HPTN 052 · PARTNER · PARTNER2 · Opposites Attract

Undetectable = Untransmittable — the evidence, plainly.

Last reviewed: September 2026

Educational information only — not medical advice. Talk to your healthcare provider about your specific situation.

U=U is not a slogan someone invented to make people living with HIV feel better. It is the plain-language summary of four of the largest prevention studies ever run — roughly 129,000 documented condomless sex acts and not one linked HIV transmission when the partner living with HIV had a suppressed viral load. Here is the whole evidence base, the definitions that matter, and the places where U=U genuinely does and does not apply.

Share Facebook X WhatsApp Text Email LinkedIn Reddit Threads Bluesky
← Home

There is a specific moment a lot of us remember. Someone — a nurse, a case manager, a stranger in a support group — says out loud that if your viral load is undetectable, you cannot pass HIV to a sexual partner. And you do not believe it. Not because the science is confusing, but because you have spent months or years or decades being told the opposite, and the body does not unlearn fear on the first hearing.

So this page is not a reassurance. It is a receipt. Below is the actual evidence base behind Undetectable = Untransmittable — the four studies, the enrollment numbers, the number of sex acts recorded, the number of transmissions found, the statistical bounds, and the exact language the CDC, the National Institutes of Health, and the World Health Organization now use. Read it once. Then read it again on a bad day.

Quick answer: Someone living with HIV who is on effective treatment with a sustained undetectable viral load cannot sexually transmit HIV. This is the strongest evidence-based statement in HIV science. The CDC puts it this way: “If you have an undetectable viral load, you will not transmit HIV through sex. This is also known as Undetectable = Untransmittable.”7 The threshold everyone is working from is a viral load under 200 copies per milliliter,8 and most people reach it within about six months of starting treatment.8

U=U is not marketing — it is the strongest consensus of the modern HIV era

It helps to understand where the phrase came from, because the origin story is often used to dismiss it. U=U was launched by the Prevention Access Campaign as a community consensus statement in July 2016.6 That means advocates named the finding before most institutions did. It does not mean advocates invented the finding.

By the time that statement went out, the underlying research had already been running for years in multiple countries, funded largely by government agencies, published in the New England Journal of Medicine, JAMA, The Lancet, and The Lancet HIV. The consensus statement's own signatories included the principal investigators of those trials — Myron Cohen of HPTN 052, Andrew Grulich of Opposites Attract, Jens Lundgren of PARTNER, and Pietro Vernazza, who co-authored the 2008 Swiss Statement that first said any of this out loud.6

What happened next is the part worth noticing. Rather than institutions grudgingly tolerating an activist claim, the claim moved into official guidance. The National Institutes of Health published a viewpoint in JAMA in 2019 titled, without hedging, “HIV Viral Load and Transmissibility of HIV Infection: Undetectable Equals Untransmittable,” authored by Robert Eisinger, Carl Dieffenbach, and Anthony Fauci.9 The CDC now states it plainly on its treatment page.7 In July 2023, the World Health Organization released new guidance stating that people who achieve an undetectable viral level through consistent antiretroviral therapy do not transmit HIV to their sexual partners and are at low risk of passing HIV to their children.10

Three of the largest public-health bodies in the world, using the same underlying data, arriving at the same conclusion. That is not a slogan. That is a scientific consensus that happens to be short enough to fit on a T-shirt.

Why this matters beyond biology: U=U changed everything — how we date, how we disclose, how the law should treat us. It moved the conversation from “how do I protect other people from me” to “how do I stay well and stay in care.” Those are different questions, and only one of them is survivable long-term.

The landmark studies, one at a time

Four studies carry most of the weight. They were designed differently on purpose — different continents, different populations, different sexual practices, different eras of treatment — and they converged. That convergence is what makes the finding durable rather than lucky.

Study 1 · 2011 and 2016 · Randomized trial

HPTN 052 — Cohen and colleagues

The HIV Prevention Trials Network enrolled 1,763 mixed-status couples across nine countries, with 54% of couples in Africa and half of the partners living with HIV being men. Every partner living with HIV had a CD4 count between 350 and 550 cells per cubic millimeter. They were randomized either to start antiretroviral therapy immediately or to wait until CD4 counts declined or illness developed.1

Cohen MS et al., New England Journal of Medicine, 20111 and 2016.2 ClinicalTrials.gov NCT00074581.

One sentence from the 2016 HPTN 052 paper does more work than any other line in this literature: “No linked infections were observed when HIV-1 infection was stably suppressed by ART in the index participant.”2 Every linked transmission in that trial happened either before suppression was achieved or after it was lost. Not one happened during stable suppression. This is the hinge on which U=U turns, and it came from a randomized controlled trial — the strongest study design medicine has.

Study 2 · 2016 · Prospective observational

PARTNER — Rodger and colleagues, JAMA

HPTN 052 answered whether treatment prevents transmission. PARTNER asked a blunter question: what happens when mixed-status couples have condomless sex, on purpose, over years, while the partner living with HIV is suppressed? Seventy-five clinical sites across 14 European countries enrolled 1,166 couples; 888 of them contributed 1,238 eligible couple-years of follow-up, with a median of 1.3 years each. Eligibility required an HIV-1 RNA level below 200 copies/mL and reported condomless sex.3

Rodger AJ et al., “Sexual Activity Without Condoms and Risk of HIV Transmission in Serodifferent Couples When the HIV-Positive Partner Is Using Suppressive Antiretroviral Therapy,” JAMA, July 12, 2016.3

Those eleven unlinked infections deserve a moment, because they are frequently misread. They are not evidence that U=U failed. They are evidence that phylogenetic analysis works — investigators could tell, at the level of viral genetics, that the virus did not come from inside the relationship. The couples were having condomless sex with each other and, in some cases, with other people. HIV arrived from the other people. It did not arrive from the suppressed partner.

Study 3 · 2019 · Prospective observational

PARTNER2 — Rodger and colleagues, The Lancet

PARTNER's first phase had relatively fewer follow-up years among gay male couples, so critics argued the data on anal sex specifically were thinner. PARTNER2 was designed to close exactly that gap. It enrolled 972 gay couples; 782 contributed 1,593 eligible couple-years, with a median follow-up of 2.0 years per couple (IQR 1.1–3.5).4

Rodger AJ et al., The Lancet, June 15, 2019.4

Study 4 · 2018 · Prospective observational

Opposites Attract — Bavinton and colleagues, The Lancet HIV

Opposites Attract took the same question outside Europe: 13 clinics in Australia plus one each in Brazil and Thailand. Of 358 male couples enrolled, 343 were followed for 588.4 couple-years. Suppression was defined as under 200 copies/mL.5

Bavinton BR et al., The Lancet HIV, August 2018.5

That PrEP-excluded analysis matters. It is the most common good-faith objection to Opposites Attract — that PrEP was doing the work. The investigators ran the numbers with PrEP periods removed and still found zero linked transmissions across more than twelve thousand condomless acts.

The numbers, added up honestly

You will see “over 100,000 sex acts, zero transmissions” quoted a lot. It is true, and it is worth being precise about how it adds up, because being sloppy with these figures gives skeptics an opening they do not deserve.

The honest arithmetic: PARTNER2 is the extension of PARTNER's gay-couple follow-up, so you cannot simply add PARTNER's full 58,000 acts to PARTNER2's 76,088 — you would double-count roughly 22,000 of them. What you can add without inflating anything is PARTNER's heterosexual arm (approximately 36,000 condomless acts),3 PARTNER2's gay-couple total (76,088),4 and Opposites Attract (16,800).5

Roughly 129,000 condomless sex acts. Zero linked HIV transmissions.

Approximately 36,000 heterosexual condomless acts in PARTNER,3 76,088 condomless anal sex acts in PARTNER2,4 and 16,800 in Opposites Attract5 — vaginal and anal, insertive and receptive, across Europe, Australia, Brazil, and Thailand. Twenty-nine HIV acquisitions occurred among HIV-negative partners in those studies. Phylogenetic analysis linked none of them to the study partner living with HIV. Layered on top: HPTN 052's randomized finding that no linked infections occurred during stable viral suppression.2

What does zero mean statistically? No study can prove a risk is exactly 0.000. What studies do is put a ceiling on how large the risk could plausibly be. PARTNER2's ceiling was 0.23 transmissions per 100 couple-years.4 PARTNER's was 0.30 overall.3 Those ceilings are the outer edge of what the data can rule out — and the observed number, in every study, at every site, was zero.

The Prevention Access Campaign chose its wording carefully for this reason. The consensus statement describes the risk as “negligible to non-existent,” and then defines the term rather than leaving it vague: “Negligible is defined as: so small or unimportant as to be not worth considering; insignificant.”6 The World Health Organization landed in the same place, describing an undetectable result as carrying zero risk of sexual transmission.11

People living with HIV who have an undetectable viral load have zero risk of transmitting HIV to their sexual partners. — Dr Lara Vojnov, World Health Organization Diagnostics Advisor, presenting WHO's updated viral-suppression guidance at the 12th IAS Conference on HIV Science, Brisbane, July 2023, as reported by aidsmap.11

What “undetectable” actually means — and the 200 number

“Undetectable” sounds absolute, but it is a lab result, and lab results depend on the machine. This trips people up constantly, so here is the distinction laid out.

Two different thresholds, both real

Modern viral load assays can often detect HIV RNA down to 20 or even fewer copies per milliliter. When your result comes back as “undetectable,” what it usually means is “below the lowest level this particular test can measure” — which might be <20, <40, or <50 depending on the lab. That number will move around between labs, and it does not mean your treatment changed.

Separately, there is the public-health threshold for viral suppression: under 200 copies per milliliter. HIV.gov defines viral suppression as 200 copies of HIV per milliliter of blood.8 The Prevention Access Campaign consensus statement uses the same number, defining undetectable as under 200 copies/mL, which it notes is also the measurement for viral suppression.6 PARTNER, PARTNER2, and Opposites Attract all set their eligibility at under 200 copies/mL.35

This is the single most reassuring thing on this page for anyone whose result came back as 43 or 120 instead of “undetectable.” The trials that produced the U=U evidence did not require you to be below 50. They required you to be below 200. If you are under 200 and staying under 200, you are inside the population these studies were built from.

WHO's three categories

In 2023, WHO formalized a three-tier framework in a policy brief on viral suppression, and it is a useful mental model.10 aidsmap described it as a traffic light.11

Why does WHO use 1,000 when U.S. guidance uses 200? Largely because of test availability. In many parts of the world, viral load monitoring runs on point-of-care platforms or dried blood spot testing, where the reliable clinical threshold sits nearer 1,000 copies/mL.11 The framework is designed for the tests people actually have access to, not the tests we wish everyone had. WHO's own figures show why access is the fight: at the end of 2022, 29.8 million of 39 million people living with HIV globally were on treatment (76%), and 71% of those had suppressed viral loads — but among children, suppression was only 46%.10

How long it takes to become undetectable

Modern antiretroviral therapy works fast. Many people see a steep drop in viral load within the first few weeks. But there is a difference between “my number dropped” and “I can rely on U=U,” and it is worth being clear about the gap.

HIV.gov's guidance is that almost everyone who takes HIV medicine as prescribed can achieve an undetectable viral load, usually within six months after starting treatment.8 The Prevention Access Campaign consensus statement puts the same point in its own frame: risk is negligible to non-existent for someone on treatment with an undetectable viral load “for at least 6 months,” and notes that “depending on the drugs employed it may take as long as six months for the viral load to become undetectable.”6

So the practical rule of thumb most clinicians and advocates use: undetectable, sustained for six months, with ongoing adherence. Not “one undetectable result.” One result is a snapshot. U=U is about a pattern.

That six-month figure is not a punishment or a probation period. It exists because it takes time to confirm that a regimen is working for your body, that you can take it consistently, and that the suppression is stable rather than momentary. If you are three months into treatment with a great number, you are doing well and you are on track. You are also allowed to wait until you and your provider are confident before you change how you approach sex.

Once you are suppressed, monitoring continues. WHO's recommended schedule is a routine viral load test at six months after starting treatment, again at twelve months, then yearly.10 In U.S. care you will often be tested more frequently. Those tests are how you know U=U still applies to you — which is a much better relationship with a lab result than dread.

Viral blips — why one detectable result does not undo U=U

This section exists because of a specific, common panic: you have been undetectable for two years, and then a result comes back at 78 copies. Your stomach drops. You start doing arithmetic about every sexual encounter in the last three months.

Take a breath. What you are most likely looking at is a blip.

What a blip is

Clinical guidelines define a blip as a newly detectable HIV RNA level, typically in the 50 to 500 copies/mL range, in someone previously suppressed, which returns below the level of quantification on prompt repeat testing.15 The key word is repeat — a blip is only definable retrospectively, after the follow-up test comes back down.

What causes them? Often nothing you did. A 2006 analysis in the Journal of Antimicrobial Chemotherapy by Lee, Kieffer, Siliciano, and Nettles concluded that most blips represent random biological and statistical variation around a mean below 50 copies/mL, or artefacts of laboratory processing — and that they were not typically associated with resistance mutations, nor with virological or clinical failure.16 Clinical guidelines add that an acute concurrent illness or a recent vaccination can also cause a transient rise.15

What to do

Where does this leave U=U? The evidence base speaks to sustained suppression — HPTN 052's finding was about infection being “stably suppressed.”2 A single blip inside a long pattern of suppression does not rewrite your history or make you a danger to anyone. What it should trigger is a repeat test and a conversation, not a shame spiral. We go deeper into this on our companion page on viral blips.

Adherence — the part U=U actually depends on

Here is the honest caveat that belongs in every U=U conversation: the equation runs on treatment that is actually in your body. U=U is not a property of having HIV. It is a property of being on effective, sustained treatment.

The consensus statement says this itself, in the same paragraph as the good news: “Continued and reliable HIV suppression requires selection of appropriate agents and excellent adherence to treatment.”6 That is not a threat. It is the mechanism.

What it does not mean is that one missed dose flips a switch. Modern regimens are built with higher barriers to resistance and longer forgiveness windows than the drugs of the 1990s, and a single late or missed dose is a common human event, not a catastrophe. What genuinely erodes suppression is a pattern — weeks of interruption, running out of medication, losing insurance, a stretch of housing instability or active depression where nothing gets taken. Those are the situations that need support, and they are situations, not character flaws.

If daily pills are the obstacle, the pill is no longer the only option.

Long-acting injectable treatment

Long-acting injectable cabotegravir plus rilpivirine is approved as a complete regimen — given monthly or every two months — for people twelve and older who have sustained virologic suppression (HIV-1 RNA under 50 copies/mL) on a stable regimen, with no history of treatment failure and no known or suspected resistance to either drug.20 HIV.gov notes plainly that long-acting injections given every two months are available.8

Two things to understand about it. First, it is a switch option, not a starting option — the federal treatment panel recommends it to replace an existing oral regimen in people who have had sustained viral suppression for at least three months, with no documented or suspected resistance to either drug, no active hepatitis B infection unless separately treated, not pregnant or actively planning pregnancy, and not taking medications with significant interactions.20 Second, it means that for some people, the number of days per year they have to think about HIV medication drops from 365 to six.

If adherence is hard right now, that is a care problem, not a moral one. Ryan White clinics, community health centers, and AIDS Drug Assistance Programs exist precisely for the gaps — cost, transportation, housing, mental health, substance use, immigration fear. Tell your care team what is actually happening. Suppression is achievable from almost any starting point, but not in secret. Start with finding care or our page for people newly diagnosed.

Pregnancy, birth, and the 2023 shift on breastfeeding

U=U is usually discussed as a sexual-transmission fact, and that is the strongest claim. But the same underlying biology has transformed pregnancy and infant feeding, and this is one of the areas where guidance has changed most recently — which means a lot of what people were told even a few years ago is now out of date.

Pregnancy and birth

NIH's patient-facing guidance is direct: when taken as prescribed throughout pregnancy, childbirth, and breastfeeding, antiretroviral therapy reduces the likelihood of perinatal transmission to less than 1 percent, and having an undetectable viral load during pregnancy and throughout breastfeeding lowers that risk to less than 1 percent.14 For people planning to conceive, treatment should be started before pregnancy where possible.14

HIV.gov describes the same pathway — treatment during pregnancy, labor, and delivery, plus four to six weeks of antiretroviral medication for the infant, reducing risk to 1% or less.8 Cesarean delivery is not recommended solely for HIV prevention when the viral load is at or below 1,000 copies/mL near delivery.14 WHO's 2023 guidance likewise states that people who achieve an undetectable viral load through consistent treatment are at low risk of passing HIV to their children during pregnancy or delivery.10

Breastfeeding — what changed on January 31, 2023

For decades, U.S. guidance told people living with HIV not to breastfeed, full stop. That changed. On January 31, 2023, the NIH Office of AIDS Research announced an update to the federal perinatal clinical guidelines that reframed infant feeding around shared decision-making.12

The substance of the update:

The guidelines were developed by the federal perinatal panel with community input, and the update explicitly recognized that restricting breastfeeding can deepen existing inequities.12 CDC's clinician guidance now calls for patient-centered, evidence-based counseling on infant feeding options to allow for shared decision-making, beginning before pregnancy or as early as possible during it, and continuing after delivery.13

Two things to hold together here. Breastfeeding while undetectable is a supported choice with a risk under 1%. It is not the same claim as sexual transmission, where the number is zero. Being precise about that difference is how we keep U=U credible — and the honest framing is not “less certain, therefore forbidden,” it is “low risk, real options, your decision, made with your clinician.” If you choose to breastfeed, providers are advised to emphasize sustained adherence and to offer close follow-up.13

What U=U does not cover — STIs, and the honest edges

U=U is a specific claim about HIV. Overstating it into a general claim about sexual health does nobody any favors, so here are the boundaries.

Other sexually transmitted infections

Maintaining an undetectable viral load does not protect you or your partners from other STIs.8 Gonorrhea, chlamydia, syphilis, herpes, mpox, and hepatitis all transmit on their own terms. The Prevention Access Campaign consensus statement makes the same point: an undetectable viral load only prevents HIV transmission to sexual partners, and condoms also help prevent HIV transmission as well as other STIs and pregnancy.6

This is not a caveat that weakens U=U. It is a reason condoms and regular STI screening remain genuinely useful tools, chosen for what they actually do rather than out of fear of a transmission that will not happen. And syphilis in particular is worth naming — rates have risen substantially in recent years, it is easily treated when caught, and it is found by testing, not by feeling fine.

Shared injection equipment

This is the genuine data gap. HIV.gov states it about as plainly as a federal agency can: “There is not enough data to know whether having a suppressed or undetectable viral load prevents HIV transmission through sharing needles, syringes, or other injection drug equipment… It very likely reduces risk, but it is unknown by how much.”8

Read that carefully, because it is often misquoted in both directions. It does not say suppression fails to protect during injection. It says nobody has measured it, so nobody can put a number on it. Sterile equipment through a syringe services program remains the intervention with actual evidence behind it. And when U=U was formally recognized in prosecutorial guidance in England and Wales in 2023, that recognition was explicitly limited to sexual transmission and did not extend to needle-sharing — the same evidentiary line.19

And what U=U was never meant to imply

U=U describes what happens when someone is on effective treatment. It says nothing about the worth of people who are not. Anyone who cannot get to care, cannot afford medication, is between insurance plans, is in active crisis, or simply has not reached suppression yet is not a failure and is not dangerous by nature. They are a person whose care system has not caught up to them. That distinction matters enormously in the legal section below.

Disclosure — what changes, and what does not

U=U rearranged the ethics of disclosure, and the rearrangement is more interesting than “you don't have to tell anyone.”

For a long time, disclosure was framed as a warning. You told someone because you posed a risk to them, and telling was how you discharged that duty. If you are undetectable and sustained, that framing collapses — there is no sexual risk to warn about.7 What remains is something closer to ordinary intimacy: you tell people about your life because you want them to know you, at the pace you choose, when you judge it safe.

That reframing is liberating and complicated at once. Some people find that U=U makes disclosure easier — it comes with good news attached, and a fact you can point to. Others find that having a scientific answer raises the expectation that they must always be prepared to teach a seminar on phylogenetic linkage in the middle of a first date. Both are real. Neither is an obligation.

The legal layer is separate

Here is the part that has to be said clearly: the science and the law are not the same thing, and being undetectable does not automatically protect you legally.

Many HIV criminalization statutes in the United States were written in the late 1980s and early 1990s and turn on disclosure or exposure rather than on transmission or actual risk. Some have been reformed; many have not. Whether U=U is relevant in your jurisdiction — as a defense, as mitigation, or not at all — depends on that jurisdiction's specific statute and case law.

Some places have moved. In April 2023, prosecutorial guidance in England and Wales was updated to recognize U=U, with cases no longer to be taken to court where a person has an undetectable viral load and is aware of U=U.19 That is real progress, and it is also one country's guidance, limited to sexual transmission.19

Nothing on this page is legal advice. If you are facing a disclosure-related legal question, talk to a lawyer with HIV-specific experience. Our page on HIV criminalization covers the landscape in more detail, and stigma and disclosure covers the human side.

Myths, answered directly

“U=U only really works for straight couples.”

Backwards, in fact. PARTNER2 was built specifically to nail down condomless anal sex among gay male couples, and produced 76,088 documented condomless anal sex acts with zero linked transmissions.4 Opposites Attract added 16,800 more among male couples across Australia, Brazil, and Thailand, again with zero linked transmissions.5 HPTN 052 covered predominantly heterosexual couples across nine countries.1 Every configuration is covered. HIV.gov states that a person with an undetectable viral load will not transmit HIV to HIV-negative partners through vaginal, anal, or oral sex.8

“U=U only counts if you use condoms too.”

The studies would not exist if that were true. PARTNER, PARTNER2, and Opposites Attract all required reported condomless sex for follow-up to count as eligible.345 Condomless sex is the exposure that was being measured. Condoms remain a good choice for STI prevention and contraception6 — they are simply not required for U=U to hold.

“One missed dose breaks U=U.”

No. Suppression is about a sustained pattern, and the finding from HPTN 052 concerned infection being stably suppressed.2 A single late dose does not undo months or years of suppression. Sustained interruption is a different matter, which is why adherence support is part of care rather than an afterthought.6

“A detectable result means I transmitted HIV to someone.”

A single detectable result in the 50 to 500 range after a period of suppression is most often a blip — random biological variation or a lab artefact — and the guidance is to repeat the test rather than draw conclusions.1516

“U=U applies to needle-sharing too.”

Not established. The federal position is that there is not enough data to know, that it very likely reduces risk, and that by how much is unknown.8 Sterile equipment is the evidence-backed answer there.

“U=U means HIV is cured.”

It does not. Suppression is achieved and maintained by ongoing treatment; stopping treatment allows viral load to rise again. What has changed is the trajectory — the international expert consensus on HIV science in criminal law describes antiretroviral therapy as having made HIV a chronic, manageable health condition.17

U=U in the courtroom — power and peril

U=U has become one of the most powerful pieces of evidence available to HIV criminalization reform. It has also created a trap that advocates saw coming and named out loud. Both halves matter.

The scientific ammunition

In July 2018, twenty leading HIV scientists — including Françoise Barré-Sinoussi, Salim Abdool Karim, Linda-Gail Bekker, Chris Beyrer, Julio Montaner, Andrew Grulich, and Mona Loutfy — published an expert consensus statement in the Journal of the International AIDS Society on the science of HIV in the context of criminal law.17 Two of its findings do enormous work in court.

First, on risk: “Current evidence suggests the possibility of HIV transmission during a single episode of sex, biting or spitting ranges from no possibility to low possibility.”17 Second, on the forensic evidence prosecutions frequently rely on: “phylogenetic analysis alone cannot prove beyond reasonable doubt that one person infected another although it can be used to exonerate a defendant.”17

The Sero Project, which works with people living with HIV facing prosecution, hosts and promotes that statement precisely because prosecutions are not always guided by the best available scientific and medical evidence.17

The trap — a viral underclass

In July 2017, a coalition led by the Center for HIV Law and Policy launched a Consensus Statement on HIV “Treatment as Prevention” in Criminal Law Reform, in response to uncertainty about how to bring modern treatment science into reform efforts.18 Original endorsers included CHLP, Prevention Access Campaign/U=U, The Counter Narrative Project, Treatment Action Group, Women With A Vision, the National Association of Criminal Defense Lawyers, the National Center for Transgender Equality, the National LGBTQ Task Force, PFLAG, and Housing Works.18

Their warning: any reform that uses viral detectability as a litmus test for criminal liability leaves the majority of people living with HIV behind and creates a viral underclass whose increased risk of prosecution effectively receives a current-day seal of approval.18 Fewer than half of all people living with HIV in the United States have sustained viral suppression.18

And suppression is not randomly distributed. The coalition's January 2018 Call to Action cited data showing sustained viral suppression at 40.8% among Black people living with HIV, 50.1% among Hispanic people living with HIV, and 56.3% among white people living with HIV.18 A legal standard built on viral load therefore imports every existing disparity in health-care access directly into criminal exposure.

Florida sits at this exact intersection. The Call to Action notes that the South carries the highest regional HIV burden in the United States, has the greatest number of uninsured people, and contains nine of the ten states with the highest incarceration rates — Florida among them.18 That combination is why Florida advocates cannot treat U=U purely as a personal-liberation message. Statewide, it has to be paired with the fight for treatment access, Ryan White capacity, and reform that does not condition dignity on a lab result.

So the position most HIV justice organizations have landed on is: use U=U to demolish the false premise that people living with HIV are inherently dangerous — and refuse to let it become the new line separating the prosecutable from the protected. As Charles King of Housing Works framed it when the CHLP statement launched, it would be a travesty for U=U to be used to create a criminal class of people who are not durably virally suppressed.18

The consensus statement itself

The Prevention Access Campaign's U=U consensus statement was issued in July 2016 and remains the reference text for the campaign.6 It is short. It is worth reading in the original rather than in paraphrase, because the precision is the point.

The core paragraph: “People living with HIV on ART with an undetectable viral load in their blood have a negligible risk of sexual transmission of HIV. Depending on the drugs employed it may take as long as six months for the viral load to become undetectable. Continued and reliable HIV suppression requires selection of appropriate agents and excellent adherence to treatment.”6

Then the definitional footwork that keeps the statement scientifically defensible: negligible defined as “so small or unimportant as to be not worth considering; insignificant”; undetectable defined as under 200 copies/mL, which is also the measurement for viral suppression; and the explicit note that an undetectable viral load only prevents HIV transmission to sexual partners, with condoms still relevant for other STIs and pregnancy.6

As of the endorsement list updated in 2021, the statement had been endorsed by more than 1,025 organizations across 102 countries — including NASTAD, the International AIDS Society, the British HIV Association, AIDS United, GMHC, the Elton John AIDS Foundation, and numerous state and city health departments — alongside individual endorsements from the principal investigators of the trials described above.6

Layer that against where the institutions have arrived independently: the CDC's flat statement that an undetectable viral load means you will not transmit HIV through sex;7 the NIH viewpoint in JAMA using “Undetectable Equals Untransmittable” in its own title;9 HIV.gov's summary that no HIV transmissions were observed when the partner living with HIV was virally suppressed;8 and WHO's 2023 guidance and policy brief.10

A community campaign, four major trials, and the three biggest public-health institutions on earth all saying the same sentence. There is very little in medicine that arrives this thoroughly triangulated.

What to do with this

Evidence only helps if it changes something. Here is what this page is actually for.

If you are newly diagnosed

Start treatment. Ask when your first viral load check will be, and what number would count as suppressed at your lab. Expect to reach undetectable within roughly six months, and understand that once you are sustained there, you cannot pass HIV to a sexual partner.8 You may need to hear that several times before it lands. That is normal.

If you are already undetectable

Know your own numbers — your last result, your lab's detection limit, how long you have been suppressed. Those three facts are the whole argument, and having them ready makes conversations with partners, family, and occasionally clinicians dramatically shorter.

If you are in a mixed-status relationship

The trials described above were built out of couples exactly like yours, and they enrolled tens of thousands of condomless sex acts to answer your specific question.345 PrEP is still available if the HIV-negative partner wants an additional layer for their own peace of mind — a third of HIV-negative partners in Opposites Attract used it5 — and choosing it is not a statement of distrust. See our PrEP page for how that works.

If you are planning a pregnancy

Bring it up before conceiving if you can, since starting treatment before pregnancy is preferred.14 Ask specifically about infant feeding, and know before you walk in that the 2023 federal update supports your choice to breastfeed if virally suppressed or to replacement feed, and states that involving Child Protective Services over infant feeding choices is inappropriate.12

If you are an advocate

Carry both halves. U=U dismantles the premise behind stigma and behind a great deal of bad law. It must not become the wall that separates suppressed people from everyone else.18 The full sentence is: people living with HIV on effective treatment cannot transmit HIV sexually, and everyone deserves the treatment access that makes that possible.

The bottom line. Four studies. Roughly 129,000 documented condomless sex acts. Zero linked HIV transmissions when the partner living with HIV was virally suppressed.345 A randomized trial finding that no linked infections occurred during stable suppression.2 The CDC, NIH, and WHO all saying it in their own words.7910 You are not a risk to the people you love. You are a person on treatment, and the science is on your side.

Related pages

References & Sources

Peer-reviewed randomized and prospective cohort studies (HPTN 052, PARTNER, PARTNER2, Opposites Attract), federal and international guidance (CDC, HIV.gov, NIH, WHO), U.S. perinatal and monitoring clinical guidelines, and the primary policy statements from the Prevention Access Campaign, the Center for HIV Law and Policy, and the international expert panel on HIV science in criminal law.

  1. Cohen MS, Chen YQ, McCauley M, et al. Prevention of HIV-1 infection with early antiretroviral therapy. New England Journal of Medicine. 2011;365(6):493–505. HPTN 052 randomized controlled trial in 1,763 mixed-status couples across nine countries: 28 virologically linked transmissions, only one in the early-therapy group; hazard ratio 0.04 (95% CI 0.01–0.27). ClinicalTrials.gov NCT00074581.
  2. Cohen MS, Chen YQ, McCauley M, et al. Antiretroviral Therapy for the Prevention of HIV-1 Transmission. New England Journal of Medicine. 2016;375(9):830–839. Final HPTN 052 results across 10,031 index person-years: 46 linked infections (3 early-ART, 43 delayed-ART), 93% risk reduction with early therapy (HR 0.07; 95% CI 0.02–0.22), and the finding that no linked infections were observed when HIV-1 infection was stably suppressed by ART in the index participant.
  3. Rodger AJ, Cambiano V, Bruun T, et al. Sexual Activity Without Condoms and Risk of HIV Transmission in Serodifferent Couples When the HIV-Positive Partner Is Using Suppressive Antiretroviral Therapy. JAMA. 2016;316(2):171–181. PARTNER study across 75 sites in 14 European countries: 888 couples contributing 1,238 eligible couple-years, approximately 22,000 condomless acts among gay male couples and 36,000 among heterosexual couples, eleven HIV acquisitions and zero phylogenetically linked transmissions; upper 95% confidence limit 0.30 per 100 couple-years.
  4. Rodger AJ, Cambiano V, Bruun T, et al. Risk of HIV transmission through condomless sex in serodifferent gay couples with the HIV-positive partner taking suppressive antiretroviral therapy (PARTNER): final results of a multicentre, prospective, observational study. The Lancet. 2019;393(10189):2428–2438. PARTNER2: 782 gay couples contributing 1,593 eligible couple-years and 76,088 reported episodes of condomless anal sex; fifteen HIV acquisitions, zero linked; upper 95% confidence limit 0.23 per 100 couple-years; authors state findings support the U=U message.
  5. Bavinton BR, Pinto AN, Phanuphak N, et al. Viral suppression and HIV transmission in serodiscordant male couples: an international, prospective, observational, cohort study. The Lancet HIV. 2018;5(8):e438–e447. Opposites Attract across 13 Australian clinics plus sites in Brazil and Thailand: 343 couples over 588.4 couple-years, 16,800 condomless anal sex acts, three HIV acquisitions and none phylogenetically linked; 12,447 acts over 232.2 couple-years in the suppressed/no-PrEP analysis.
  6. Prevention Access Campaign — Undetectable = Untransmittable (U=U) Consensus Statement. The July 2016 consensus statement: negligible-to-non-existent sexual transmission risk with an undetectable viral load sustained at least six months, the definition of “negligible,” the under-200 copies/mL threshold, the note that U=U covers sexual transmission only, and endorsement by more than 1,025 organizations in 102 countries plus the trial principal investigators. Text and endorsement list verified via the archived snapshot.
  7. U.S. Centers for Disease Control and Prevention — Treating HIV. CDC's plain statement that with an undetectable viral load you will not transmit HIV through sex, also known as Undetectable = Untransmittable.
  8. HIV.gov — HIV Treatment as Prevention. Federal summary: viral suppression defined as 200 copies/mL, undetectable usually achieved within six months of starting treatment, no transmission through vaginal, anal, or oral sex when suppressed, long-acting injections every two months, perinatal risk reduced to 1% or less, no protection against other STIs, and the explicit statement that there is not enough data on shared injection equipment.
  9. Eisinger RW, Dieffenbach CW, Fauci AS. HIV Viral Load and Transmissibility of HIV Infection: Undetectable Equals Untransmittable. JAMA. 2019;321(5):451–452. NIH leadership viewpoint adopting the U=U framing in the peer-reviewed literature.
  10. World Health Organization — The role of HIV viral suppression in improving individual health and reducing transmission (policy brief, 2023). WHO's three categories (unsuppressed above 1,000 copies/mL; suppressed but detected at or below 1,000; undetectable), with zero sexual transmission risk when undetectable and negligible risk when suppressed, plus the routine monitoring schedule at six months, twelve months, then yearly. Announced in the WHO news release New WHO guidance on HIV viral suppression and scientific updates released at IAS 2023 (23 July 2023), which also reports that 29.8 million of 39 million people living with HIV were on treatment at the end of 2022, 71% of them virally suppressed, but only 46% of children.
  11. aidsmap — An amber light: the World Health Organization's position when an HIV viral load is suppressed but not undetectable. IAS 2023 conference reporting including Dr Lara Vojnov's statement that people living with HIV with an undetectable viral load have zero risk of transmitting HIV to their sexual partners, the traffic-light framing of the three categories, the role of the PARTNER and Opposites Attract zero-transmission findings as the scientific underpinning of U=U, and WHO's recommendation of enhanced adherence support with a repeat test in three months for suppressed-but-detectable results.
  12. NIH Office of AIDS Research — Update to Clinical Guidelines for Infant Feeding Supports Shared Decision Making: Clarifying Breastfeeding Guidance for People with HIV (January 31, 2023). The federal perinatal guideline shift: zero postnatal risk with replacement feeding, less than 1% but not zero with sustained undetectable viral load while breastfeeding, clinicians should support either choice, and the statement that it is inappropriate to engage Child Protective Services in response to infant feeding choices.
  13. U.S. Centers for Disease Control and Prevention — Reducing Transmission of HIV Through Breast Milk (clinician guidance). CDC's figure of less than 1% but not zero transmission risk through breastfeeding with sustained undetectable viral load, plus its recommendation for patient-centered, evidence-based infant feeding counseling and shared decision-making, and close follow-up for those who breastfeed.
  14. NIH HIVinfo — Preventing Perinatal Transmission of HIV During Pregnancy and Childbirth. Antiretroviral therapy taken as prescribed through pregnancy, childbirth, and breastfeeding reduces perinatal transmission to less than 1%; treatment should ideally begin before conception; cesarean delivery is not recommended solely for HIV prevention when viral load is at or below 1,000 copies/mL near delivery.
  15. New York State Department of Health AIDS Institute Clinical Guidelines Program (Johns Hopkins University) — Virologic and Immunologic Monitoring in HIV Care. Definition of a viral blip (newly detectable 50–500 copies/mL returning below quantification on prompt repeat), retesting within four weeks (two weeks if 500 copies/mL or above), the instruction that treatment should not be changed on a single elevation, and the elevated failure risk associated with consecutive blips and sustained low-level viremia.
  16. Lee PK, Kieffer TL, Siliciano RF, Nettles RE. HIV-1 viral load blips are of limited clinical significance. Journal of Antimicrobial Chemotherapy. 2006;57(5):803–805. Analysis concluding that most blips reflect random biological and statistical variation around a mean below 50 copies/mL or laboratory processing artefacts, and are not typically associated with resistance mutations or with virological or clinical failure.
  17. Barré-Sinoussi F, Abdool Karim SS, Albert J, et al. Expert consensus statement on the science of HIV in the context of criminal law. Journal of the International AIDS Society. 2018;21(7):e25161. Twenty-scientist consensus: per-act possibility of transmission through a single episode of sex, biting, or spitting ranges from no possibility to low possibility; phylogenetic analysis alone cannot prove beyond reasonable doubt that one person infected another though it can exonerate a defendant; antiretroviral therapy has made HIV a chronic, manageable health condition. Hosted and promoted for people facing prosecution by the Sero Project.
  18. Center for HIV Law and Policy et al. — Call to Action for Racial Justice in HIV Criminal Law Reform (January 2018), with the Consensus Statement on HIV “Treatment as Prevention” in Criminal Law Reform (launched July 2017). The viral-underclass warning against using viral detectability as a litmus test for criminal liability; fewer than half of people living with HIV in the U.S. have sustained viral suppression; sustained suppression of 40.8% among Black, 50.1% among Hispanic, and 56.3% among white people living with HIV; the South's HIV burden, uninsured rate, and nine of the ten highest-incarceration states including Florida; and the full endorser list. The Charles King quotation on not creating a criminal class of people who are not durably suppressed is reported by the HIV Justice Network.
  19. aidsmap — U=U acknowledged in prosecutors' guidance in England and Wales (April 2023). Updated Crown Prosecution Service guidance recognizing that cases should no longer be taken to court where a person has an undetectable viral load and is aware of U=U, with the limitation that this applies to sexual transmission and not to shared injection equipment.
  20. Panel on Antiretroviral Guidelines for Adults and Adolescents (NIH/HHS) — Optimizing Antiretroviral Therapy in the Setting of Virologic Suppression. Federal guidance on long-acting injectable cabotegravir plus rilpivirine: approved monthly or every two months for people 12 and older with sustained virologic suppression (HIV-1 RNA under 50 copies/mL) on a stable regimen with no treatment failure history and no known or suspected resistance, and the Panel's switch criteria including at least three months of sustained suppression. Indication language confirmed in the FDA CABENUVA prescribing information.