Aim: To assess whether patients with misdiagnoses of chronic pancreatitis (CP), followed at an early stage by a diagnosis of pancreatic cancer (PCr), present different epidemiological characteristics from patients suffering either from CP alone or from CP with late degeneration to PCr. Methods: We arbitrarily subdivided our patient series into three groups: (1) 12 CP who developed PCr within 4 years after onset of symptoms; (2) 12 CP developing PCr after the 4th year, and (3) 701 CP with no subsequent development of PCr. The variables studied were age, sex, drinking and smoking habits, tumor localization, and presence of intraductal calcifications and diabetes mellitus at the time of diagnosis of CP. Results: There were no significant differences between CP and ‘late’ PCr in any of the study variables considered. As compared with the CP group, the ‘early’ PCr cases were older (58.7 vs. 40.7 years; p < 0.0001), with a lower proportion of males (58 vs. 88%; p < 0.01), smaller proportions of both smokers (42 vs. 88%; p < 0.0001) and subjects drinking more than 40 g of alcohol/day (42 vs. 86%; p < 0.0001), and a greater incidence of non-insulin-dependent diabetics at the time of diagnosis of CP (25 vs. 3.7%; p < 0.012). As compared with the ‘late’ PCr group, the malignancies in the ‘early’ PCr cases were more often located in the head of the pancreas (100 vs. 50%; p < 0.01). Multivariate logistic regression analysis selected age over 50 (odds ratio OR 13.5, 95% confidence interval CI 2.79–65.5; p < 0.001), smoking habits (OR 0.14, 95% CI 0.04–0.49; p < 0.002), and non-insulin-dependent diabetes (OR 5.91, 95% CI 1.20–29.1; p < 0.028) as variables identifying subjects with ‘early’ PCr. Conclusions: A high suspicion of a pancreatic tumor is necessary when CP is diagnosed in a patient with atypical epidemiological characteristics for this condition, possibly female, aged over 50, who is not a smoker or drinker, and suffers from non-insulin-dependent diabetes.