To evaluate the risk of another cancer among persons who initially developed Kaposi's sarcoma, the authors used data from the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute for the years 1973-1990. In persons under 70 years of age, 4,946 cases of Kaposi's sarcoma were observed during the period 1980-1990 (6,217 person-years of follow-up). On the basis of rates seen during the period prior to the epidemic of acquired immunodeficiency syndrome (AIDS), 169 cases were expected. Therefore, cases of Kaposi's sarcoma in this group were assumed to be AIDS-related, while cases occurring in older persons or during the 1970s were assumed to be non-AIDS-related. Rates were compared with the numbers of cases expected overall and by site on the basis of age-, sex-, and calendar year-specific rates from the SEER data. Among the 4,946 persons with AIDS-related Kaposi's sarcoma, the risk of developing non-Hodgkin's lymphoma through 1990 was increased 198-fold (95% confidence interval 169-232). However, the risk of all other cancers was only marginally increased (1.5-fold; 95% confidence interval 0.95-2.3), a risk that was probably biased upward because of ascertainment and misclassification. Among 491 persons with non-AIDS-related Kaposi's sarcoma, the relative risk of all cancers, including non-Hodgkin's lymphoma, was 0.9 (upper 95% confidence limit 1.2), and the risk of non-Hodgkin's lymphoma alone was 0.6 (upper 95% confidence limit 3.3). As of 1990, the risk of having another cancer following Kaposi's sarcoma was increased only in persons infected with human immunodeficiency virus, who were at high risk of non-Hodgkin's lymphoma but probably not of other cancers as a whole.