Gemcitabine (2',2'difluoro-2'deoxycytidine, dFdC) is a synthetic antimetabolite of the cellular pyrimidine nucleotide metabolism. In a first series of in vitro experiments, the drug showed a strong effect on the proliferation and colony formation of the human androgen-sensitive tumor cell line LNCaP and the androgen-insensitive cell lines PC-3 and DU-145. Maximal inhibition occurred at a dFdC concentration as low as 30 nM. In contrast to the cell lines which were derived from metastatic lesions of prostate cancer patients, no inhibitory effects were found in normal primary prostatic epithelial cells at concentrations up to 100 nM. The effect of gemcitabine was reversed by co-administration of 10-100 pM of its natural analogue deoxycytidine. In view of a future clinical application of this anti-tumor drug in advanced prostatic carcinoma, we have compared the effect of gemcitabine on prostatic tumor cells with that on bone marrow granulopoietic-macrophagic progenitor cells, because neutropenia is a common side effect of gemcitabine treatment. The time course of action on the two kinds of cells was markedly different. Colony formation of tumor cells was inhibited by two thirds at a gemcitabine concentration of about 3.5 nM. The same effect on granulopoietic-macrophagic progenitor cells required a concentration of 9 nM. Co-administration of deoxycytidine to gemcitabine-treated tumor cell cultures completely antagonized the effect of gemcitabine whereas addition of deoxycytidine after 48 hr of gemcitabine treatment could not prevent gemcitabine action on the tumor cells. In contrast, more than half of the granulopoietic-macrophagic progenitor cells could still be rescued by deoxycytidine administration after 48 hr. These findings and the marked difference in the susceptibility of neoplastic and normal prostatic cells suggest that gemcitabine is a promising substance which should be further evaluated as to its efficacy in the treatment of advanced prostatic carcinoma.