Whether long-term use of vitamin K antagonists (VKAs) might affect the incidence of cancer is a longstanding hypothesis. We conducted a population-based study including all cancer- and thromboembolism-free patients of our health area; study groups were defined according to chronic anticoagulant use to VKA-exposed and control groups. Cancer incidence and cancer-related and overall mortality was assessed in both groups. 76 008 patients (3231 VKA-exposed and 72 777 control subjects) were followed-up for 8.2 (± 3.2) years. After adjusting for age, sex, and time-to-event, the hazard ratio of newly diagnosed cancer in the exposed group was 0.88 (95% confidence interval [95% CI] 0.80-0.98; P < .015). VKA-exposed patients were less likely to develop prostate cancer, 0.69 (95% CI 0.50-0.97; P = .008). The adjusted hazard ratio for cancer-related and overall mortality was 1.07 (95% CI 0.92-1.24) and 1.12 (95% CI 1.05-1.19), respectively. These results support the hypothesis that anticoagulation might have a protective effect on cancer development, especially prostate cancer.