[Show abstract][Hide abstract] ABSTRACT: Survival in patients with laryngeal cancer has not increased remarkably within the last years. It is presumed that a variety of factors act jointly in predicting survival after diagnosis: tumour stage, tumour site, treatment approaches, age and comorbidities. The aim of this German clinical multi-centre study is to present results from multivariate analysis. A retrospective cohort study was conducted in four hospitals in South-West Germany. Incident cases with laryngeal squamous cell carcinoma were included for the years 1998 to 2004, resulting in a population sample of 594 patients. Multivariate regression analysis was performed using the Cox proportional hazards model. Patients were followed up for 64.1months on average. Overall 5-year survival was 66% (95% confidence interval (CI): 62-70%). The strongest risk factors in multivariate analysis were age at first diagnosis (hazard ratio (HR): 1.5; 95% CI: 1.5-1.7 per each additional 10years), tumour stage, and the development of recurrences (HR 3.1; 95% CI: 2.3-4.2) or second primary carcinomas (HR 2.1; 95% CI: 1.4-3.1). A somewhat weaker effect was shown for patients with comorbidities (using Charlson's comorbidity index). The choice of treatment did not strongly affect survival when adjusting for other factors, possibly because the optimal treatment approach was applied for the specific constitution and requirements of each patient. For future research it would be desirable to study the effect of treatment on quality of life in multivariate analysis as well as other modifiable risk factors as smoking and drinking reduction or cessation after diagnosis.