We designed a randomized clinical trial to test whether the addition of three cycles of chemotherapy during standard radiation therapy would improve disease-free survival in patients with stages III and IV (i.e., advanced oropharynx carcinoma). A total of 226 patients have been entered in a phase III multicenter, randomized trial comparing radiotherapy alone (arm A) with radiotherapy with concomitant chemotherapy (arm B). Radiotherapy was identical in the two arms, delivering, with conventional fractionation, 70 Gy in 35 fractions. In arm B, patients received during the period of radiotherapy three cycles of a 4-day regimen containing carboplatin (70 mg/m(2) per day) and 5-fluorouracil (600 mg/m(2) per day) by continuous infusion. The two arms were equally balanced with regard to age, sex, stage, performance status, histology, and primary tumor site. Radiotherapy compliance was similar in the two arms with respect to total dose, treatment duration, and treatment interruption. The rate of grades 3 and 4 mucositis was statistically significantly higher in arm B (71%; 95% confidence interval [CI] = 54%-85%) than in arm A (39%; 95% CI = 29%-56%). Skin toxicity was not different between the two arms. Hematologic toxicity was higher in arm B as measured by neutrophil count and hemoglobin level. Three-year overall actuarial survival and disease-free survival rates were, respectively, 51% (95% CI = 39%-68%) versus 31% (95% CI = 18%-49%) and 42% (95% CI = 30%-57%) versus 20% (95% CI = 10%-33%) for patients treated with combined modality versus radiation therapy alone (P =.02 and.04, respectively). The locoregional control rate was improved in arm B (66%; 95% CI = 51%-78%) versus arm A (42%; 95% CI = 31%-56%). The statistically significant improvement in overall survival that was obtained supports the use of concomitant chemotherapy as an adjunct to radiotherapy in the management of carcinoma of the oropharynx.