Preclinical data is consistent with the concept that the timing of chemotherapy during radiant heat-whole body hyperthermia (WBH) should affect therapeutic index. In order to test this hypothesis, a controlled clinical investigation was initiated. Patients received carboplatin (CBDCA) on an early or late schedule with respect to achieving target temperature (i.e. 41.8 degrees C) in alternating treatment cycles. The first cycle was randomized between patients regarding the early or late schedule for two planned sets per patient (i.e. four cycles). Ifosfamide, etoposide and granulocyte colony stimulating factor were delivered during all cycles with a standardized schedule. A total of 53 cycles involving 17 patients were analyzed. Detailed toxicity evaluation (i.e. delay in therapy secondary to thrombocytopenia, need for platelet transfusions, and days of hospitalization) taken collectively demonstrated a statistically and clinically significant advantage to delivering CBDCA 10 min after target temperature, during the plateau phase of WBH.