In order to determine the value of induction chemotherapy (CT) for inoperable head and neck cancer, the authors conducted a retrospective study. Fifty-five patients were treated with CT and radiotherapy (R-T)(CT+RT group). This group was compared with a group of 54 patients treated RT alone (RT alone group). The CT regimen used were CF (cis-platine+5-FU), CVB (cyclophos-phamide+vincristine+bleomycin), CAP (cyclophosphamide+adriamycin+prednisolone) or PVBM(cis-platine+vincristine+bleomycin+methotrexate). Toxicity from induction chemo-therapy was minimal, and toxicity was limited primarily to nausea and vomiting, mucositis and myelosuppression. The complete response (CR) rate to CT was $14.5\%$ and the partial response (PR) rate was $47.3\%$ for an overall major response rate of $61.8\%$. The major response rate at the completion of loco-regional therapy was $87.3\%$(48/55) with 32 CR ($58.2\%$) and 16 PR ($29.1\%$) for CT-RT group and $81.5\%$(44/55) with 27 CR ($50.0\%$) and 17 PR ($31.5\%$) for RT alone group (p=0.57). Median follow-up of CT-RT group was 17 months and 11 months for RT alone group. Median survival was 30 months for CT-RT group and 24 months for RT alone group (p=0.3). The overall survival rate at 2 years, 3 years and 5 years, respectively was $60.9\%,;48.6\%;and;42.5\%$, for CT-RT group, and $54.9\%,;49.9\%;and;49.9\%$ for RT alone group (p=0.33). Comparision between patients in both groups, stratified by overall stage, T and N stage, site, and pathology, all failed to show any significant difference in survival rates. We conclude that this retrospective study failed to demonstrate an advantage for induction chemotherapy in inoperable head and neck cancer.