Although the biochemical and cytokinetic effects of thymidine (TdR) have been extensively studied in vitro, other parameters may be encountered in vivo that could alter its therapeutic efficacy. L1210 cells in culture are sensitive to the growth inhibitory effects of TdR, while TdR infusions in mice bearing L1210 ascites fail to prolong survival. Our results demonstrate that the TdR levels achieved in the serum and ascites fluid during sc and iv infusions range between 10(-4) and 10(-3) M. These levels are sufficient to inhibit the growth of bone marrow and intestinal mucosa cells, as evidenced by 32P incorporation studies. However, these TdR levels fail to exhibit the growth of L1210 ascites, despite the fact that these cells are sensitive to similar concentrations in vitro. The absence of growth inhibition in the L1210 ascites is shown to be due to increased endogenous CdR levels (approximately 2 X 10(-5) M) in the ascites fluid. The concomitant serum CdR levels were at least ten times lower than those achieved in the ascites. This factor probably accounts for the differential effect of TdR on normal cells as compared in vivo with agents such as TdR or even antimetabolites, such as cytarabine, may be modulated by increased CdR levels in the local tumor milieu. These studies are relevant in that TdR is currently being used in a variety of clinical protocols using the nucleoside alone at high doses or in combination with methotrexate, 5-FU, or cytarabine.