Growth of human adenocarcinomas of the colon and rectum in immunoincompetent mice has allowed for a greater understanding of the interaction of 5-fluorouracil, its metabolism, and mechanism(s) of cytotoxicity under conditions of tumor growth in situ. Conversely, this agent has proven to be a useful tool in defining metabolic characteristics in human colon adenocarcinomas. Analysis of tumor sensitivity to 5-fluorouracil (FUra),5-fluorouridine (FUrd) and 5-fluoro-2'-deoxyuridine (FdUrd) suggests that growth inhibition in vivo is related to a DNA-directed event. Resistance, de novo appears to be a consequence of relatively transient inhibition of the target enzyme thymidylate synthase (dTMP-synthase), which may be a consequence of low concentrations of 5,10-methylenetetrahydrofolate (CH2-H4PteGlu) or its polyglutamate forms within tumor cells in situ. In order to study the relationship between inhibition of dTMP-synthase and growth inhibition, mutant cells deficient in their ability to salvage dThd have been selected, and grown as xenografts. Data suggest that transient inhibition of dTMP-synthase and not dThd salvage is responsible for resistance de novo, and that prolonged inhibition of dTMP-synthase would be a lethal event in vivo. This would predict that a cell lacking dTMP-synthase activity would not be tumorigenic. This has been tested directly by selecting clones of GC3 colon adenocarcinoma cells deficient in dTMP-synthase (TS-) activity. Preliminary data indicate that each of 3 TS- clones is tumorigenic in athymic nude mice. The importance of dTMP-synthase as a target for drug development is discussed with respect to these findings.