The quinazoline folate analogue raltitrexed (ZD1694; Tomudex) and the camptothecin derivative irinotecan are two new agents showing clinical activity against colorectal cancer. To identify the optimal conditions to achieve synergistic cytotoxicity before the clinical development of their combination, we explored the interactions between ZD1694 and the active metabolite of irinotecan, 7-ethyl-10-hydroxycamptothecin (SN-38), in vitro. Cytotoxicity was evaluated with a clonogenic assay using the human colon cancer cell line HCT-8. Different schedules of administration and different dose ratios of the two agents were compared and evaluated for synergism, additivity, or antagonism with a quantitative method based on the median-effect principle of Chou and Talalay (T. C. Chou and P. Talalay, Adv. Enzyme Regul., 22: 27-55, 1984). Sequential short-term (1 and 4-h) exposures to SN-38 followed by ZD1694 resulted in synergistic cytotoxicity at broad dose-effect ranges. At a high level of cell kill, the synergism was greater when either equiactive doses of the two agents or higher relative doses of ZD1694 were used. A 24-h interval between exposure to SN-38 and ZD1694 significantly enhanced the magnitude of the synergy (P = 0.001). The opposite sequence or simultaneous exposures produced significantly less potentiation or nearly additive interactions (P = 0.0006 and P < 0.0001). The synergism was completely lost under conditions of more prolonged drug exposure (24-h continuous exposure). In conclusion, in this in vitro model of human colon cancer, ZD1694 and SN-38 produced synergistic cytotoxicity. Our findings support the clinical use of this combination and provide a rationale for a clinical trial using sequential short-term exposures to equiactive doses of SN-38 and ZD1694 administered sequentially with a 24-h interval.