They are seen by many AIDS experts as a key to stopping new infections, and Nomvula Mathiso can explain why.

"My boyfriend says he doesn't want to use a condom, and so now I can use the gel and just keep quiet about it," said Ms. Mathiso, 30, one of the volunteers testing the microbicide.

Her sister died of AIDS in 2001, at 33, and Ms. Mathiso knows many more people who have the disease. (One in three people in Soweto lives with HIV-AIDS.) In sub-Saharan Africa, cultural and social factors often mean women have limited or no say over sexual choices such as condom use. As a consequence, they make up three-quarters of those infected with HIV-AIDS in some areas.

A microbicide would offer them an HIV-prevention method that they control, and they wouldn't even need to tell their partners about it.

"This increases the choices controlled by women," said Sibongile Walaza, a researcher supervising the trial.

"Women say, 'If he's drunk or he just doesn't want to use a condom, there's nothing I can do about it,' and so it will be great to have some choice."

Results from this study will likely come in late 2008 at the earliest. Dr. Walaza said the researchers are hopeful, based on lab tests with the product, that the volunteers who use the gel will show a lower rate of HIV infection than those who had the placebo in their applicators.

Participants such as Ms. Mathiso are given free condoms and extensive counselling about safer sex, and reminded that PRO 2000 might not work, but the reality of human behaviour is that researchers can still expect a certain level of HIV infection in the test cohort. What the researchers are hoping to see is roughly half as many infections in those who got the gel as those who didn't.

The gel, developed by the U.S. company Indevus Pharmaceuticals Inc., is expected to protect women against herpes, gonorrhea and chlamydia in addition to HIV. But the drug will still allow conception to occur, which will be a key factor in making the product palatable to many African women.

The trial, which has also begun in Uganda and will soon expand into Tanzania and Zambia, is being co-ordinated by Britain's Medical Research Council. The size of this trial suggests real optimism about whether the drug will work.

"There's clearly been a tipping point," said Zeda Rosenberg, who heads the International Partnership for Microbicides in New York, an agency working to accelerate microbicide development.

"The G8 [ the group of eight leading industrial nations, which cited the importance of microbicides in the final communiqué from its summit in Gleneagles, Scotland, in July], the global advocacy campaign, people are starting to get it. It's truly a sea change."

And while the pace of drug development is achingly slow, given the size of the crisis, she said, microbicide development has picked up real momentum lately after years in obscurity.

A major obstacle in developing such a product is the fact that it will be of use mostly to women in poor countries, which gives major pharmaceutical companies, the originators of most new drugs, very little incentive to develop it.

But this past month, Bristol-Myers Squibb Co. and Merck & Co. Inc. each licensed new drugs to the IPM to develop further, exempting the IPM from royalties and authorizing the agency to distribute any microbicide made with the drugs free to the developing world.

The drugs are a new class of anti-retrovirals (the drugs best known as treatment for HIV) called entry inhibitors, and they work by interfering with the Velcro-like attachment that allows HIV to get into a cell.

One of the drugs binds to proteins on the virus itself, impairing its ability to attach to a cell, while others provide a similar function on the cell side. By messing up the proteins the virus wants to latch on to, it shields the cell from infection.

Merck and BMS were developing the drugs with an eye to using them as a treatment for people already infected with HIV. But when researchers at Tulane University and Cornell University in the United States tested them in monkeys to see if they prevented infection, they found the drugs worked better together than alone. The IPM stepped in to say it could facilitate development of the drugs used together as a microbicide.

In a study published recently in the journal Nature, John Moore at Cornell and Ronald Veazey of the Tulane National Primate Research Center combined several drugs into a topical ointment that they applied to the vaginas of macaque monkeys to test for protection against simian HIV.

One compound protected 21 out of 28 macaques treated 30 minutes before being exposed to HIV. When the three compounds were combined, they completely blocked infection in three treated monkeys. The compounds also provided protection up to six hours after application.

Those are hugely encouraging results -- for monkeys. But there is a big jump to humans.

"What the monkey studies do is give you a sense of the direction of effect," Dr. Rosenberg said.

The number of monkeys was too small to say anything definitive, but "it's enough to say 'they look promising.' The main issue we all have as scientists and drug developers is we don't know what any of these mean until we have a study on efficacy in women."

In Soweto, Ms. Mathiso is ready to spread the word.

"I just want other women to know: If this gel works, maybe this disease will be beaten."