To determine if use of antiherpes drugs protects against the development of AIDS-associated Kaposi's sarcoma (KS), data from 935 homosexual men with AIDS from the Multicenter AIDS Cohort Study were analyzed. In nested case-control analysis, neither acyclovir use for human immunodeficiency virus infection (odds ratio [OR], 0.84; 95% confidence interval [CI], 0.56-1.26; P = .39) nor acyclovir use for any indication (OR, 1.02; 95% CI, 0.76-1.38; P = .89) was associated with a reduced risk of KS as initial AIDS diagnosis. In longitudinal analysis, acyclovir was also not protective against developing KS as a late manifestation of AIDS (after initial non-KS AIDS diagnosis). Among men with cytomegalovirus disease, ganciclovir use (relative risk [RR], 0.56; 95% CI, 0.22-1.44; P = .23) and foscarnet use (RR, 0.40; 95% CI, 0.051-3.10; P = .38) were associated (although not significantly) with a reduced risk of KS. Thus, acyclovir use does not appear to reduce the risk of KS, but further study of other antiherpes drugs such as ganciclovir and foscarnet is warranted.