Poz4Poz saves lives.

Eventually even The New York Times had to report it. HIV infection rates among gay men seem to be going down in San Francisco. Five years ago the Times put on its front page a prediction from San Francisco health authorities that we were on the verge of a "sub-Saharan" level of new infections. Oops. San Francisco has just cut its estimate of new infections almost by half. But what's really freaking out the AIDS establishment is one theory as to why this may be happening.

Serosorting. Never heard of it? We soon will. It's not a new idea. I remember a friend of mine with AIDS well over a decade ago telling me he had the answer to the HIV epidemic among gay men, something that would stop it in its tracks. "Segregation," he said. "I refuse to have sex with people who are HIV-negative." If every HIV-positive man did that, the epidemic would soon collapse, he argued. The man is still alive, thriving actually. And yes, he forsakes condoms with other HIV-positive men, and he argues that it's an incentive to keep the firewall between positives and negatives in place.

There is an obvious objection. You don't control whom you fall in love with or whom you find attractive. Neg-poz relationships can work, safe neg-poz sex can work, and the virus should not be able to dictate the rules of love. MI true. But at the same time the danger of infecting your loved one is real for serodiscordant couples, even when consistently safe. And even in casual safer sex, accidents happen. Other things being equal, it's better to find a partner with the same serostatus. If we could push the trend more in that direction, we'd undoubtedly reduce new infections. And if holding out the possibility of more intimate condom-free sex helps keep HIV within the HIV-positive world, why not use it as a plus? Same for neg-neg monogamous relationships. The reward for loving one man exclusively can be intimacy and freedom from viral fear. And both strategies could reduce infection rates.

As for casual or anonymous sex, disclosure is still vital. What seems to have happened in San Francisco is that men with HIV now form a critical mass of potential partners, disclose their status up front, and keep the virus within the boundaries of their own viral world. Internet hookups may even help--because it's less intimidating to announce your status online than to another man's face. "Poz 4 Poz" and "Neg-UB2" are slogans we need to see more of. Ditto "Poz 4 Poz" and neg-only Web sites.

Equally, knowledge is power, and getting tested is a prerequisite for serosorting. That's why we're seeing progress among the more enlightened parts of gay America--like San Francisco--but still see worrying trends among closeted men with less information and income. Yes, there's a small chance of reinfection with another viral strain for poz couples. But the risk is low and still debated, especially for people on meds, and certainly lower than when having sex with men whose status you don't know.

Serosorting is not a miracle cure, and it merits more debate. I'm just trying to be constructive here and put the idea on the table. What's undeniable is that the old universal condom code--based on terror of death--is broken, as HIV has become a manageable illness, if still one any sane person would avoid. Since the disincentive of fear won't work as well anymore, the incentive of intimacy might help stem the virus's spread. HIV segregation may be part of the answer. Perhaps it already is.