The loss of CD4 lymphocytes is known to be an important component of human immunodeficiency virus type 1 (HIV-1) pathogenesis. It remains unclear, however, whether the importance of the CD4 lymphocyte count is such that individuals who have been infected for widely different lengths of time, but for whom the CD4 lymphocyte count is the same, have the same risk of developing acquired immunodeficiency syndrome (AIDS). The authors directly addressed this question for 111 HIV-1-infected hemophiliacs who had been followed for up to 12 years from seroconversion and for whom a median of 16 CD4 lymphocyte counts had been made. Thirty-eight patients had developed AIDS by January 1, 1992. As of August 1, 1985, the time from seroconversion to AIDS ranged from 3 weeks to almost 6 years. The risk of AIDS increased with time from seroconversion (relative hazard = 1.25 per year; p = 0.09). CD4 lymphocyte count was strongly related to time from seroconversion and also to the risk of AIDS (relative hazard = 3.25 per 100/mm3; p < 0.00005). In a bivariate Cox regression model, the relative hazard for time from seroconversion fell markedly towards unity (1.05; p = 0.7), while that for CD4 lymphocyte count remained unchanged. This suggests that time from seroconversion is a relevant factor in HIV-1 infection only insofar as it provides information as to the probable degree of loss of CD4 lymphocytes. If the number of these cells is known, time from seroconversion seems to be of little additional value in assessing the risk of AIDS. This finding has implications for the clinical assessment of individual patient prognosis, for the design of epidemiologic studies of HIV-1 infection, and for our understanding of how HIV-1 causes disease.