The complex relationships among tumor cell burden, dose and schedule of chemotherapy, and efficacy were investigated in a murine osteosarcoma model in which an easily measured marker provided an accurate, dynamic estimate of host tumor cell burden. Cytoxan (200 mg/kg) produced a 93.2-99.997% tumor cell kill in animals with a pretreatment tumor burden of 0.6-3% body weight. In animals with a pretreatment tumor burden of 5.1-10.24% body weight, however, the same dose of cytoxan produced <1 order of magnitude tumor cell kill (31-71%). A schedule utilizing three doses q12 days in animals with a moderate burden of tumor (up to 5% body weight) produced a cell kill of six or more frequent orders of magnitude with some cures, an event which was more frequent with added immunostimulation. A schedule utilizing two doses q20 days in animals with a larger body burden (5-10% of body weight) was essentially ineffective. These results suggest that a small initial body burden (low stage) and an aggressive schedule of treatment are necessary for optimum results in cancer treatment. Small delays in initiation of treatment and prolonged intervals between doses can convert an effective drug to an ineffective one without the need to invoke biochemical mechanisms of resistance.