Objective: To investigate the influence of previous tumours on overall survival in patients with early glottic carcinoma. Design: Retrospective, population-based cohort study. Setting: Cancer registration area in the west Netherlands. Methods: Population-based data on previous and subsequent tumours in patients diagnosed with early glottic carcinoma (Tis and T1) in the west Netherlands between 1982 and 1993 were collected from charts and cancer registries. The impact of previous tumours on survival was tested in multivariate analysis. Main Outcome Measures: Overall survival. Results: Of 359 patients, 22 patients (6%) had a total of 23 previous tumours. Previous tumours had an independent impact on overall survival (hazard ratio 3.4, p <= .001). Other determinants of survival were age and subsequent tumours. Nonmalignant comorbidity was of borderline significance. Conclusions: Previous tumors have an impact on survival in patients with early glottic carcinoma and should be considered as an additional factor in counseling and prognostication.