Pouring cold water on one more anti-HIV weapon, researchers have said circumcision does not necessarily protect gay men from HIV.
Whatever the previous results, more definitive evidence is required to show that circumcision does have such beneficial effects, says the Journal of the American Medical Association.
African trials have shown circumcision cuts the likelihood of female to male HIV transmission by up to 60%.
The US team, from the Centers for Disease Control and Prevention in Atlanta examined data on 53,567 men who have sex with men, of which 52% were circumcised.
HIV levels among those who were circumcised were lower - but not significantly so.
However, there was evidence that circumcision may have had a protective effect in studies carried out before the introduction of highly active antiretroviral therapy (HAART) in 1996.
The researchers said one possible explanation was that the advent of these more effective HIV drugs had encouraged higher levels of risk taking sexual behaviour, wiping out any marginal protective effect from circumcision.
Alternatively, the drugs may have helped lower the risk of transmission to the point where circumcision had no further benefit.
A third possibility is that there may have been a smaller proportion of men in the pre-HAART trials who primarily engaged in receptive anal sex, which carries the greatest risk for HIV infection among gay men.
Michael Carter, of the HIV information service NAM, said: "It's my sense that there has been a tempering of the excitement about circumcision, and researchers are now favouring a 'combination prevention' approach.
"Circumcision may have a place in this in some settings, but so too do good sexual health, consistent condom use, and there's real excitement and debate about the role of HIV treatment in prevention."
Will Nutland, of the HIV charity Terrence Higgins Trust, said: "This research adds weight to the evidence that circumcision isn't an effective method of HIV prevention for men who have sex with men.
"The majority of HIV infections in men who have sex with men are as a result of receptive anal intercourse and circumcision would make no difference in these cases.
"Rather than encouraging gay men to be circumcised, investment in prevention in the UK should focus on targeted education programmes, condom provision and easy access to testing."