Although the incidence of classic Kaposi's sarcoma (CKS) has been investigated, its occurrence following a primary neoplasm and its association with this first neoplasm need to be determined. We analyzed a series of 124 patients with a secondary CKS (8.4% of a total of 1485 incident cases) which occurred between 1961 and 1992 in the Jewish Israeli population. Data on first neoplasms and subsequent Kaposi's sarcoma were retrieved from the Israel Cancer Registry. Acquired-immune-deficiency-syndrome-related Kaposi's sarcomas were excluded from the case series. Four controls were randomly selected for each CKS case among all Cancer Registry cases free from a second neoplasm at the time of diagnosis of the CKS in the case, and matched on gender, year of birth and year of diagnosis of the first neoplasm. The average time lapse between first neoplasm and secondary CKS was 4.5 years, being shorter for prostate cancer and for hematopoietic malignancies. As compared with Israel-born Jews, the risk of a subsequent CKS was significantly increased in immigrants [odds ratio (OR) 3.0]; this risk was particularly high in immigrants from the former Soviet Union (OR 9.4) and Poland (OR 7.0). There was no clear trend with age at immigration; however, low age at immigration and a short length of stay in Israel endowed a higher risk of developing a secondary CKS, markedly among patients suffering from solid tumors as the first primary. There was an excess of secondary CKS following a non-Hodgkin's lymphoma (OR 5.3), a Hodgkin's lymphoma (OR 7.5), a leukemia (OR 5.3) or a breast cancer (OR 2.2). Cancer patients with a first primary in the lung, colon, stomach, larynx, liver, pancreas or kidney showed secondary CKS less frequently. Despite the lack of control of therapy for the first neoplasm, development of secondary CKS seems to be mediated by mechanisms similar to those for hematopoietic neoplasms and selected nonhematopoietic neoplasms, such as breast cancer. The trend toward increased risk after a short time lapse and the difference in risk among immigrants indicate that genetic susceptibility is part of the complex interplay between cellular proliferation and control systems.