The effects of conventional vs protected housing, and those caused by the periodic application of a psychological stressor (rotational stress, spatial disorientation) on mice kept in a protected housing, with spontaneous tumor metastasis have been determined in mice implanted with Lewis lung carcinoma as a function of tumor inoculum size and response to treatment with cyclophosphamide and razoxane. With a reduced inoculum size, tumor takes do not occur in mice kept in the protected housing, but do occur with spatial disorientation. With a larger inoculum size, tumor takes occur in all untreated mice, and the weight of spontaneous lung metastasis is significantly increased by spatial disorientation. For mice in protected housing, cyclophosphamide results in the absence of macroscopically detectable tumors in all of the treated mice, whereas the use of spatial disorientation abolishes this therapeutic action. The antimetastatic effects of razoxane are also reduced by rotational stress. These results indicate that housing conditions and a psychological stressor can control tumor takes and metastasis formation. They also indicate that host's antitumor resistance effectors, which are susceptible to neuroendocrine modulation by environmental and psychological stressors, participate to determine the effectiveness of the treatment with a cytotoxic (cyclophosphamide) and antimetastatic (razoxane) antitumor drug.