BEIJING, March 21-- Two field tests to find an AIDS vaccine and seven other trials of similarly designed AIDS vaccines have either been stopped or put off indefinitely after it was found the most promising contender not only did not protect people from the virus but may actually have put them at increased risk of becoming infected.

The results of the trials, which enrolled volunteers on four continents, have spurred intense scientific inquiry and unprecedented soul-searching as researchers try to make sense of what happened and assess whether they should have seen it coming.

Numerous experts are questioning both the scientific premises and the overall strategy of the nearly 500 million U.S. dollars in AIDS vaccine research funded annually by the U.S. government.

"This is on the same level of catastrophe as the Challenger disaster" that destroyed a NASA space shuttle, said Robert Gallo, co-discoverer of the human immunodeficiency virus (HIV), which causes AIDS, and head of the Institute for Human Virology in Baltimore.

The recently closed studies, STEP and Phambili, used the same vaccine -- made from a common respiratory virus called adenovirus type 5 that had been crippled and then loaded with fragments of HIV. Both studies were halted when it became clear the STEP study was futile and possibly harmful.

The results of the Phambili vaccine trial, which was conducted in South Africa, were revealed last month and only worsened the gloom. Although the number of new HIV infections in that study was far smaller than in STEP -- and too few to draw firm conclusions from -- those results, too, hinted at a trend toward harm among vaccine recipients.

Many researchers are questioning the scientific premises on which all those studies were based and are wondering, along with AIDS activists, what effect this near-worst-case scenario might have on tests of future vaccines.

The working hypothesis for what went wrong is that the vaccine somehow primed the immune system to be more susceptible to HIV infection -- a scenario neither foreseen nor suggested by previous studies.