The hypothesis that human immunodeficiency virus (HIV) causes AIDS was advanced in 1984, based only on circumstantial evidence. To this date, the primary evidence are correlations between the presence of antibody against HIV and AIDS. But these correlations are biased by proponents of the HIV hypothesis in favour of HIV. They ignore HIV-free AIDS and they base correlations on selected studies because there are no national HIV-AIDS statistics. The HIV-AIDS hypothesis has made the following predictions: (I) AIDS would ‘explode’ from the original risk groups into the general population via sexual transmission of HIV (2) Health care workers would contract AIDS from their patients, scientists from propagating HIV, and prostitutes from their clients. (3) The 150 chimpanzees that have been experimentally inoculated with HIV, and the 15,000 American hemophiliacs who have been iatrogenically inoculated before 1984, would develop AIDS. (4) Antiviral immunity and vaccines would protect against AIDS. (5) HIV would cause AIDS by killing T-cells. (6) AIDS would occur only in people infected by HIV. But none of these predictions proved to be correct. Recent studies show that HIV is a passenger virus instead of the cause of AIDS: (1) AIDS occurs at unpredictable intervals after infection; (2) HIV may be active, passive, or totally absent from otherwise identical AIDS cases. Indeed, AIDS does not meet one of the classical criteriy of infectious disease: (1) Equal distribution between the sexes; (2) disease following infection within days or weeks, the time microbes take to become either immunogenic or pathogenic or both; (3) the presence of a common active microbe. Therefore it is proposed that American and European AIDS is caused by the long-term consumption of recreational drugs and the anti-HIV drug AZT This hypothesis is testable and provides a rational basis for AIDS control.