The association between calcium intake and gastrointestinal cancer mortality was investigated in a 28-year follow-up study. Data were obtained from a general health examination in 1953-1954 among Dutch civil servants and their spouses, aged 40 to 65 years. Information from 2,591 participants was used for this study. Risk analyses were performed using logistic regression models with the highest quintile of calcium intake as reference. No statistically significant relation between calcium intake and gastrointestinal cancer mortality is observed, although the risk estimate for women with the lowest calcium intake is substantial. Odds ratios (OR) were adjusted for age, energy intake and dietary fiber. After inclusion of other potential confounders to the models (body-mass index and smoking habits), the results hardly differed. Both men and women who died of colorectal cancer had a lower mean calcium intake compared to the rest of the population. For women this was statistically significant. Our results suggest that a low calcium intake may be related to gastrointestinal cancer mortality among women of this study population.