The aim of this study was to assess the potential prognostic value of tobacco and alcohol consumption in 161 consecutive patients with intra-oral squamous cell carcinoma. The patients were included in a prospective clinico-experimental and epidemiological study to secure valid data on tobacco and alcohol consumption. Univariate analysis showed that patients with an alcohol consumption above the median had a significantly (P = 0.03) poorer prognosis than other patients, with 5-year cause-specific survivals of 54 +/- 6 and 33 +/- 6%, respectively. Similarly, patients with a tobacco consumption above the median had a worse prognosis than other patients, with 5-year cause-specific survivals of 55 +/- 6 and 39 +/- 6%, respectively. This difference was on the borderline of significance, P = 0.056. Tobacco and alcohol consumption were correlated and each of them correlated with sex, males having a higher consumption than females. T-stage, N-stage, clinical stage, tumour size and sex all had a significant prognostic impact. To elucidate whether tobacco and alcohol consumption had an independent prognostic value, a multivariate analysis by means of the Cox proportional hazards regression analysis was performed. This analysis showed that clinical stage (P = 2 x 10(-5)), tumour size (P = 0.007) and tobacco consumption (P = 0.046), but not alcohol consumption, had significant influences on prognosis. Thus, smoking cessation programmes seem warranted both from a prophylactic and a therapeutic point of view.