Irinotecan [7-ethyl-10-(4-[1-piperidino]-1-piperidino)-carbonyloxy-camptothec in] administered i.v. in two courses, each course consisting of administration every day for 5 days [(dx5)2] on days 1-5 and 8-12, has demonstrated significant activity against advanced human tumor xenografts derived from colon adenocarcinomas and several childhood cancers. To build on this therapy, we have evaluated the combination of irinotecan given on this schedule with 5-fluorouracil given on days 1, 7, and 14 with or without leucovorin [(dx5)3 i.v.] against colon tumors, or combined with etoposide administered (dx5)2 i.v. either 2 h before or 2 h after irinotecan for treatment of colon tumors and rhabdomyosarcomas. A combination of 5-fluorouracil at 75% and irinotecan at 50% of their respective maximum tolerated doses when administered as single agents on this schedule gave acceptable toxicity. Against colon adenocarcinoma xenografts, 5-fluorouracil did not enhance the response rate compared with that obtained with the optimum dose of irinotecan given as a single agent. Against GC3/TK- xenografts, which lack thymidine kinase and cannot salvage thymidine to circumvent the inhibition of thymidylate synthase, the addition of leucovorin to the combination increased the complete response rate from 10 to >90%, whereas the response rates for the optimal doses of irinotecan or 5-fluorouracil, as single agents, were 30 and <10%, respectively. Etoposide d x 5 i.v. for two or three courses or (d x 5)3 p.o. did not cause objective regression of any colon tumors. In contrast, three of five rhabdomyosarcoma lines demonstrated a high frequency of partial regressions or complete regressions when treated (d x 5)1 i.v. Repetitive courses [e.g., (d x 5)2 or (d x 5)3] i.v. or p.o. or by 4-h infusion d x3 i.v. were either equally effective or less effective. Irinotecan and etoposide were combined using the (d x 5)2 i.v. schedule for both drugs, in which irinotecan was given 2 h before or 2 h after the administration of etoposide. Each drug could be combined at only 38% of its respective maximum tolerated dose when administered as a single agent, indicating greater than additive toxicity. Toxicity was similar irrespective of the sequence of administration and was manifested by loss of weight (73% of the initial weight, nadir day 7), myelosuppression, and prolonged thrombocytopenia. The responses of colon carcinomas to the combination given in either sequence were similar to that achieved with irinotecan given alone at the same dose as used in the combination. Similarly, when etoposide was given before irinotecan, the responses of rhabdomyosarcomas were similar to those for irinotecan. However, in experiments in which etoposide was administered 2 h after each dose of irinotecan, there was significant antagonism of the antitumor activity of irinotecan.