When cyclophosphamide, 1-beta-D-arabinofuranosylcytosine, vincristine, methylprednisolone (P) and cis-diamminedichloroplatinum (CP) were administered to mice previously given injections of 4.5 or 5 million L1210 leukemia cells, the effectiveness of the 5-drug combination was influenced by the stage of the circadian system at the time of injection. By applying what we refer to as the chronobiological approach (timed treatment), in comparison with a homeostatic (time unqualified) approach, fewer deaths and less weight loss were found, as the result probably of lower drug toxicity. Despite a cure rate that ranged from 24 to 48% as a function of CP timing in the first study, the overall acute drug toxicity (ranging from 20 to 76%) was unacceptable as a treatment protocol. In a second study, by lowering dosages of all drugs but still administering the drugs in a chronobiological manner, death due to acute drug toxicity was reduced to zero while the percentage of cures ranged from 44 to 88% in animals treated with P at different circadian stages. In both studies the homeostatic approach was unsatisfactory because of overwhelming drug toxicity.