Undetectable viral load is one of the best-established facts in HIV medicine, and it is routinely explained with the wrong number. Plenty of pages, including the one this article replaces, give 20 copies per millilitre as the threshold. That figure is the sensitivity of a machine, not the point below which HIV stops being transmitted sexually. The threshold that every study behind U=U actually used is 200, and the gap between those two numbers has caused a great deal of avoidable fear.
What an undetectable viral load actually means
200 copies per millilitre, and where the number came from
Every trial and cohort that established zero sexual transmission defined suppression as fewer than 200 copies of HIV RNA per millilitre of blood. Not 20, not 50. The clinical guidelines follow suit: viral suppression is under 200 copies/mL, while “undetectable” as a laboratory word simply means whatever the assay in use cannot see.
Why a lab report saying 35 is not bad news
Modern assays detect far below the threshold that matters. Roche COBAS reads down to 20 copies/mL, Hologic Aptima to 30, Abbott RealTime to 40, bioMérieux NucliSENS to 10. So a result can read “HIV-1 RNA detected, 35 copies/mL” for someone whose undetectable viral load is, by every meaningful definition, exactly that. Researchers at the University of Miami went as far as calling the reporting of those values without context a harmful medical practice, and proposed that laboratories either annotate them with “no risk of sexual transmission” or report anything under 200 as undetectable outright.
The evidence: four studies, zero linked transmissions
HPTN 052, the randomised trial
Cohen and colleagues randomised 1,763 serodifferent couples to early or delayed antiretroviral treatment. Of 72 genetically linked infections, three occurred in the early-treatment group against 43 in the delayed group, a 93% reduction. The three early-treatment cases occurred before suppression was achieved or after treatment failure, which is precisely the point.
PARTNER, PARTNER2 and Opposites Attract
PARTNER1 followed 888 serodifferent couples through roughly 58,000 acts of sex without condoms and recorded zero phylogenetically linked transmissions. PARTNER2 followed 782 gay male couples through 76,088 condomless acts, again with zero linked transmissions despite 15 new infections in the cohort, all traced to partners outside the couple. Opposites Attract added 12,447 acts among 343 male couples, also zero.
That is well over 140,000 acts of condomless sex with an undetectable viral load on one side, and not a single transmission attributable to the partner. Few claims in sexual health rest on evidence that clean.
Below the threshold, and just above it
Blips
A blip is a single detectable reading under 200 copies/mL in someone otherwise suppressed. The clinical guidance is unambiguous: people whose viral load remains undetectable or suppressed do not transmit HIV sexually, even during a temporary blip, provided they stay on treatment. There is no evidence that a value between the assay’s floor and 200 predicts anything bad at all.
The amber light between 200 and 1000
A systematic review of eight studies covering 7,762 serodifferent couples across 25 countries found no transmissions below 200 copies/mL and only two possible transmissions anywhere under 1,000, both in the 600 to 1,000 range with long gaps between the viral load test and the infection. On that basis the WHO now describes zero risk for an undetectable viral load and almost zero or negligible risk for one that is suppressed but detectable. Useful to know, and not a reason to relax about a rising result.
Durably undetectable: the six-month rule
Why six months
The guidelines define durably undetectable as suppression maintained for at least six months. Most people reach an undetectable viral load within six months of starting treatment taken as prescribed. The six-month window is not a safety buffer invented for reassurance, it is what establishes that suppression is stable rather than a single good result.
What breaks it
Stopping treatment breaks it. So do long interruptions, and the pattern where doses slip during travel, weekends or chemsex, which is the same failure mode that undoes oral PrEP, discussed in the article on injectable PrEP. U=U holds while treatment is taken. It is a statement about people on treatment, not a permanent property of a person.
What U=U does not cover
Every other STI
An undetectable viral load does nothing against syphilis, gonorrhoea, chlamydia, hepatitis or mpox. None of them care about HIV RNA. Regular screening and vaccination remain exactly as necessary as before, and barriers keep the role described in how condoms actually fail.
Needles, and what was never studied
This is the honest limit. The guidelines state plainly that research has not established that an undetectable viral load prevents transmission through shared needles or needlestick injuries. Nobody has run that study. Anywhere blood and sharps are part of play, U=U is not the answer, and the absence of evidence should be read as absence rather than reassurance.
What this changes in practice
Serodifference is not a risk category
Someone with a durably undetectable viral load poses no sexual HIV risk to a partner. That is not a hedge or a probability, it is what more than 140,000 recorded acts show. A great deal of the caution that circulates in kink and fetish spaces about serodifference is not caution, it is stigma running on outdated information, and it pushes people away from testing and treatment, which is the one thing that genuinely raises risk for everyone.
Does the negative partner still need PrEP?
Inside a couple where one partner has a durably undetectable viral load and neither has sex outside it, PrEP adds nothing to what suppression already does. That calculation changes the moment there are other partners, casual play, or any doubt about how consistently treatment is being taken, and it is a decision to make with a prescriber rather than by reading a table. The point is that PrEP and U=U answer different questions: one protects a person across every partner they have, the other describes one specific partner who cannot transmit.
When to talk to a doctor
A rising viral load, a result above 200, or any stretch of missed doses is a conversation with a prescriber, not a forum, and it is worth having quickly rather than after the next test. Anyone who has had a real exposure risk should look at the site’s page on what to do after a transmission risk, where the 72-hour window is the thing that matters. And anyone who has been told their undetectable viral load is 35 copies and left the appointment worried has been let down by the report, not by the treatment.
