Antiviral therapy directed against human immunodeficiency virus type 1 (HIV-1) is now approaching its 10-year anniversary. The initial clinical trial of zidovudine was completed in September 1986, and, with expedited review, zidovudine was approved by the Food and Drug Administration (FDA) early in 1987 (1). The ensuing years have seen the introduction of many other nucleoside analogues that inhibit reverse transcriptase, including (in order of their approval by the FDA) didanosine, zalcitabine, stavudine, and lamivudine. All of these drugs inhibit HIV in vitro, all are associated with a decrease in viral RNA concentrations in serum, all are associated with an increase in CD4 cell counts, and four of the five are known to reduce rates of progression to the acquired immunodeficiency syndrome (AIDS) or prolong survival, or both. The principal problems of these drugs are their limited antiviral activity, their toxicity, and their lack of a durable antiviral effect, which is at least partly explained by the development of resistance. The result is what has come to be called a “time-limited benefit,” which has provoked substantial controversy about the relative merits of early and late initiation of treatment.