A randomized, prospective trial utilizing cisplatin and fluorouracil as neoadjuvant chemotherapy in the treatment of advanced squamous cell carcinomas of the upper aerodigestive tract was initiated in January 1983. Sixty patients were stratified by site (oral cavity, 19; larynx, 14; hypopharynx, 14; oropharynx, 11; nasopharynx, one; and paranasal sinuses, one) and by stage (III, 19; IV, 41), and then randomized to receive either standard treatment (defined as preoperative irradiation followed by radical excision or irradiation alone) or adjuvant chemotherapy followed by standard treatment. An additional three patients were entered into the study, but withdrew. Chemotherapy consisted of three cycles for those patients in whom an objective tumor response was observed; nonresponders received standard treatment. Response to chemotherapy was complete in five and partial (greater than 50%) in 18 patients, for an overall response rate of 85%. The follow-up for surviving patients was a minimum of 24 months and a maximum of 44 months. Survival was compared for patients in both treatment groups according to the method of Lee and Desu. Despite excellent tumor response, actuarial survival was 70% in the standard treatment group as opposed to 56% in the experimental group. It was therefore evident that the high response rates reported in previous pilot studies do not necessarily result in improved survival in these cancers.