Are we winning or losing the battle against AIDS? It's a timely question as we close out a dramatic year filled with good news and bad.

First, there was the announcement that an experimental AIDS vaccine from Merck & Co. had proved ineffective in an advanced trial and, even worse, may have caused an increased susceptibility to HIV infection in volunteers. Then came a report by the Joint United Nations Program on HIV-AIDS and the World Health Organization that we are beginning to see a small reduction in the number of annual new HIV infections globally.

So what does this mean? How does it add up?

Unfortunately, we're still losing against AIDS, badly. But there are good reasons to think we can win, with the tool that holds the hope of eliminating, and not just curbing, the epidemic: a preventive vaccine. The failure of one product does not rule out success on that front.

Having changed their methodology, UNAIDS and the WHO now estimate that there are 33 million people living with HIV-AIDS, down from 40 million, their 2006 figure. Still, AIDS remains the world's fourth-leading cause of death, and it is No. 1 in sub-Saharan Africa. It has killed at least 23 million people.

UNAIDS and the WHO said that the trend toward fewer new infections probably reflects the natural progression of the epidemic as well as behavioral changes brought on by prevention programs. While it is impossible to sort out those factors, campaigns such as those that encourage abstinence, faithfulness to a single partner, circumcision, the use of condoms and the use of clean needles by intravenous-drug users are critical and need more attention. Still, given human nature and the cultural and economic realities in the societies hit hardest by AIDS, these programs can do only so much.

If we want to protect society as a whole and eliminate AIDS, we need a much more powerful weapon: a vaccine. No major viral epidemic has been defeated without one.

Whatever the precise dimensions of the AIDS epidemic, it is horrendous, especially in sub-Saharan Africa, which accounts for more than two-thirds of all new infections. It may comfort some to believe that condoms, circumcision and sermons will trump human nature. But if the virus were as widespread in the United States as it is in Africa, we would be engaged in a frantic medical Manhattan Project to find a vaccine.

And we would find one. The history of medical progress says so.

Twenty-four years ago, when a virus was found to be the cause of AIDS, there were only a few licensed medicines for use against any virus, and scientists thought it implausible to develop drugs against HIV. Today, we have more drugs to treat HIV than for all other viruses put together. That's what we can do when we focus the power of science.

It took 47 years after the virus responsible for polio was identified before scientists developed a vaccine for the disease. That we are free from iron lungs as well as a range of other infectious scourges is a debt we owe to the persistence and optimism of previous generations. Now it is our turn.

We have much better tools at our disposal than early vaccinologists had, but we're still in the early stages. To date, only two experimental AIDS vaccines have completed efficacy testing. We had high hopes for the Merck candidate, but more study is needed to find out why it failed and what implications that may have for the 30 or so experimental AIDS vaccines in trials.

That an AIDS vaccine is possible is strongly suggested by the fact that most people's immune systems hold HIV in check for years before they develop AIDS. A small number of infected people seem never to develop the disease. We also know, from studies in non-human primates, our closest relatives, that vaccines can protect from infection with SIV, HIV's cousin in monkeys.

We need to build on these insights as quickly as possible. We can't afford the alternative, in financial or human terms. In September, UNAIDS estimated that it would cost at least $45 billion a year by 2015 to achieve the Group of Eight's goal of providing a comprehensive package of AIDS prevention, treatment and care to everyone who needs it in the developing world. That estimate will decline along with the revised epidemiological figures. But even with the drop in infection rates, AIDS spending, barring a revolution in giving, would have to go from one-tenth to one-quarter of all foreign aid to achieve the G-8 goal. What would get cut? Funds for basic healthcare? Agriculture? Education? Clean water?

Focusing on the long-term goal of creating an AIDS vaccine is undeniably hard. The world is spending some $900 million a year on the effort, up from less than $200 million a decade ago. It is still a minimal investment, but it is one that will ultimately prove cost-effective: A preventive vaccine is the only intervention that could ultimately eliminate the need for all others.

Seth Berkley is chief executive and president of the International AIDS Vaccine Initiative. He wrote this commentary for The Washington Post.

Make a difference. Donate to The Advertiser Christmas Fund.