We are moving rapidly beyond a "black box" understanding of the pathogenesis of HIV. The sites of virus replication, the molecular regulation of virus production in the host, and the dynamics between productive virus infection and immunological and clinical events are areas of intense study using powerful new tools. The quantitation of virus load and genetic characterization of replicating virus has important implications for the development and evaluation of drugs and treatment strategies for HIV. As new compounds are introduced, their ability to reduce virus load in vivo has become a primary consideration in the decision to initiate large efficacy trials and may soon be used, in combination with other markers, in the licensing of new agents. In parallel, rapid molecular evaluation of virus from patients, targeting those who break through drug-induced suppression, provides an explanation for the failure of drugs to sustain an effect on virus load. This approach has compressed the process of drug evaluation and set the stage for the evaluation of complex combinations and sequences of drugs to maintain suppression of virus and prevent the development of drug resistance. The most controversial question for the next few years is whether the measurement of virus load or detection of drug resistance can be incorporated into the practice of medicine and the management of individual patients. There is evidence that changes in virus load are the most proximate markers of drug response and that detection of resistance mutations can predict clinical and immunological decline. However, the window of time between a change in load or the development of drug resistance and a decline in CD4 cells is relatively short. With dideoxynucleoside therapies, a CD4 cell decline follows a rise in virus load or development of resistance within 3-6 months. In early studies with protease inhibitors and nonnucleoside reverse transcriptase inhibitors, the development of resistance and a return to baseline of virus load may occur within 2-3 months, mirrored by a fall in CD4 cells. The challenge to investigators is how to best use these new tools to determine whether changes or additions in therapy, initiated on the basis of virological measurements, result in more effective management of disease.