Aims/hypothesis Proinsulin is possibly associated with cancer through activation of insulin receptor isoform A. We sought to investigate the associations between proinsulin and 20 year cancer mortality rates. Methods The study was performed within the Hoorn Study, a population-based study of glucose metabolism in individuals aged 50-75 years in the Dutch population. Fasting proinsulin levels were measured twice by a doubleantibody radioimmunoassay. Participants were continuously followed to register mortality; causes of death were derived from medical records. Cox survival analyses were performed to assess the 20 year risk of death from cancer in relation to proinsulin. All analyses were adjusted for age and sex, with additional adjustments for traditional risk factors. The effect modification of glucose metabolism and sex was tested.Results Proinsulin levels were measured in 438 individuals (41% normal glucose tolerance, 35.7% impaired glucose metabolism, 23.3% type 2 diabetes). Of these participants, 53 died from cancer. After adjustment for age and sex, proinsulin >16.5 pmol/l (the upper tertile) was significantly associated with a twofold risk of cancer mortality (HR 2.01, 95% CI 1.16, 3.46) compared with individuals with lower proinsulin levels. Additional adjustment for glucose metabolism, BMI and smoking did not substantially change the results (HR 1.91, 95% CI 1.04, 3.52). No interaction with glucose metabolism or sex was observed. Conclusions/interpretation Individuals with fasting proinsulin levels >16.5 pmol/l have a twofold risk of cancer mortality over a 20 year time span. These findings provide population-based evidence for the independent association between high proinsulin levels and cancer mortality rates.