We examined the effects of dosage schedule on antitumor activity in vitro and in vivo to determine the optimal administration schedule for a new nucleoside antimetabolite 1-(3-C-ethynyl-beta-D-ribo-pentofuranosyl)cytosine (ECyd, TAS-106). The cytotoxicity of TAS-106 in vitro against human tumors was evaluated at three drug exposure periods. TAS-106 exhibited fairly potent cytotoxicity even with 4 h exposure, and nearly equivalent and sufficiently potent cytotoxicity with 24 and 72 h exposures. These results suggest that long-term exposure to TAS-106 will not be required to achieve maximal cytotoxicity. The antitumor activity of TAS-106 in vivo was compared in nude rat models bearing human tumors on three administration schedules, once weekly, 3 times weekly, and 5 times weekly for 2 or 4 consecutive weeks. TAS-106 showed strong antitumor activity without serious toxicity on all three schedules, but the antitumor activity showed no obvious schedule-dependency in these models. When tumor-bearing nude rats were given a single i.v. dose of [(3)H]TAS-106, tumor tissue radioactivity tended to remain high for longer periods of time as compared to the radioactivity in various normal tissues. Furthermore, when the metabolism of TAS-106 in the tumor was examined, it was found that TAS-106 nucleotides (including the active metabolite, the triphosphate of TAS-106) were retained at high concentrations for prolonged periods. These pharmacodynamic features of TAS-106 may explain the strong antitumor activity without serious toxicity, observed on intermittent administration schedules, in nude rat models with human tumors. We therefore consider TAS-106 to be a promising compound which merits further investigation in patients with solid tumors.