The Head and Neck Intergroup conducted a Phase III randomized trial to determine whether the addition weekly cisplatin to daily radiation therapy (RT) would improve survival in patients with unresectable squamous cell head-and-neck carcinoma. Eligible patients were randomized to RT (70 Gy at 1.8-2 Gy/day) or to the identical RT with weekly cisplatin dosed at 20 mg/m(2). Failure-free survival (FFS) and overall survival (OS) curves were estimated with the Kaplan-Meier method and compared with the log rank test. Between 1982 and 1987, 371 patients were accrued, and 308 patients were found eligible for analysis. Median follow-up was 62 months. The median FFS was 6.5 and 7.2 months for the RT and RT + cisplatin groups, respectively (p = 0.30). The p value for the treatment difference was p = 0.096 in multivariate modeling of FFS (compared to a p = 0.30 in univariate analysis). Expected acute toxicities were significantly increased with the addition of cisplatin except for in-field RT toxicities. Late toxicities were not significantly different except for significantly more esophageal (9% vs. 3%, p = 0.03) and laryngeal (11% vs. 4%, p = 0.05) late toxicities in the RT + cisplatin group. The addition of concurrent weekly cisplatin at 20 mg/m(2) to daily radiation did not improve survival, although there was evidence of activity. Low-dose weekly cisplatin seems to have modest tumor radiosensitization but can increase the risk of late swallowing complications.