Treatment with the cytotoxic antitumor drug cyclophosphamide is highly effective in mice bearing Lewis lung carcinoma, causing the absence of macroscopically detectable tumors at necroscopy after sacrifice. When the effects of the treatment on survival are determined, a significant increase in survival time and in the proportion of long-term survivors is observed. When restraint stress is further applied, tumors develop in all of the mice treated with cyclophosphamide, and survival time and the fraction of long-term survivors are significantly reduced. Flow cytometry of splenic T-lymphocyte subsets in normal mice indicates a significant decrease in the number of CD3+, CD4+, and CD8+ subsets after treatment with cyclophosphamide and after application of restraint stress; the interaction of the two treatments is significant for CD3+ and marginally significant for CD4+ subsets. The attenuation by restraint stress which was observed for the effects of cyclophosphamide on the presence of tumors at necroscopy and for the survival of the treated mice might thus be interpreted as follows: restraint stress attenuates the immune functions of the host directed toward the weakly immunogenic tumor, an effect which, in the absence of restraint stress, interacts effectively with the cytotoxic action of cyclophosphamide toward tumor cells. The results obtained using this animal model thus indicate that experimental stress reduces the therapeutic efficacy of a cytotoxic antitumor drug; experimental and clinical implications are discussed.