5535 Background: Prospective randomized trials failed to show any survival benefits associated with the use of induction chemotherapy in nasopharyngeal carcinoma (NPC). Although cisplatin-based regimens were commonly used in these trials, the impact of cisplatin dose was not clear. METHODS Individual patient data of two published phase III trials using different cisplatin-based induction chemotherapy regimens were pooled together for analysis. A total of 365 patients with stage II-IV NPC received either cisplatin & epirubicin (EP, n = 144) or cisplatin, bleomycin & fluorouracil (PBF, n = 221) followed by radiotherapy between September 1989 and July 1994. The mean total cisplatin dose received by patients ranged from 162 mg/m2 in EP group to 232 mg/m2 in PBF group. Median follow-up time was 59 months. RESULTS The response rates at the end of induction chemotherapy were 86% in PBF group and 84% in EP group. No significant differences in tumor control and survival were observed between the two groups. Five-year overall survival rates were 63% in PBF group and 62.5% in EP group. Five-year loco-regional failure-free rates were 74% in PBF group and 65% in EP group, the difference being not significant after adjusting for stage grouping. CONCLUSIONS No significant difference in treatment outcome was observed in patients using different cisplatin-based induction chemotherapy. The absence of survival benefits in chemotherapy arms was probably not related to the dose-intensity of cisplatin used, and escalating the total cisplatin dose alone in induction chemotherapy is not likely to improve survival. No significant financial relationships to disclose.