The study of stem cell defects in man after cytoreductive therapy appears to be of particular importance since hematotoxicity represents the major limiting side effect in many instances. On the other hand, such studies are met by difficult methodological problems. In particular, bone marrow sampling in man cannot be done on a quantitative basis. Furthermore, ethical considerations restrict the sampling of serial bone marrow specimens. We have studied the reaction of hematopoiesis including stem cells in man with the available methods, in particular CFU-GM (CFU-C), CFU-E, and BFU-E in bone marrow (BM) and peripheral blood (PB) during the course of various cytostatic regimens given for adjuvant and palliative chemotherapy of human cancer. Major findings of these studies were The acute reaction of BM stem cells and differentiated BM precursor pools corresponds to the mechanisms known to be effective in animal experiments. The PB CFU-GM may be of particular importance in man in demonstrating long-term derangements. In an attempt at quantification of human BM data, an index was developed from BM spicule morphometry and from standardized BM aspirate cellularity. Stem cell defects in BM and PB were observed that did not show up in the PB counts, in particular after nitrosoureas. Indication of prolonged derangements in stem cells up to five years after chemotherapy were observed after adjuvant therapy for breast cancer. From these data, it appears advisable to study stem cell data in man for newly developed chemotherapeutic regimens in order to minimize early and late side effects on hematopoiesis.