The mortality experienced by a cohort of 653 patients with celiac disease in Edinburgh and the Lothian region has been analyzed. Mortality overall was 1.9-fold (95% confidence limits, 1.5-2.2) that of the general population (115 deaths observed, 61.8 expected; p less than 0.0001). The increased mortality was greatest within 1 yr of diagnosis of celiac disease and steadily declined over time with the excess mortality being concentrated at ages 45-54 yr in men and 55-64 yr in women. Celiac disease was mentioned on the death certificate in 33 cases but in only 10 was it given as the underlying cause of death. Of 17 deaths from lymphoproliferative diseases (0.55 expected, p less than 0.001), 8 occurred within 2 yr of diagnosis of celiac disease compared with 8 (0.37 expected, p less than 0.001) occurring greater than 5 yr after diagnosis. Esophageal cancer was certified as the cause of four deaths (0.47 expected, p less than 0.01). In men mortality from all other malignant disease was also increased (15 deaths observed; 6.4 expected, p less than 0.01), but most of these deaths occurred within 5 yr of the diagnosis of celiac disease. In contrast, there was no deficit in deaths from ischemic heart disease or stroke and the mortality rate in those diagnosed in childhood as having celiac disease was similar to the general population.