The effect of combined treatment of UFT with CDDP, especially the optimal time-dependent administration sequence, was studied in nude mice bearing nine human cancer xenografts derived from stomach, colon, lung, breast and uterocervical cancers following treatment with clinically equivalent doses and routes of administration. This combination treatment produced a higher response rate than a single treatment in 7 out of 9 cancers (78%). However, the combination timing of UFT and CDDP was assumed to be an important factor to produce a good therapeutic effect, since the combination effect was different by changing the treatment procedure, e.g., UFT followed by CDDP or CDDP followed by UFT. There was a low correlation between combination effect and the treatment procedure, origin of tissues, histology and doubling time of cancer cells or efficacy of a single agent. The strongest antitumor effect was observed by the treatment with UFT administered both prior to CDDP and after CDDP. Such treatment procedure seemed to be reasonable for cancers of unknown origin or character.