To determine behavioural, dietary and other common factors associated with new cases of gallstones, diagnosed by ultrasonography, in a prospective cohort study conducted in southern Italy. Between May 1985 and June 1986, systematic sampling from the electoral register of Castellana, a small town in southern Italy, yielded 2472 subjects who had had their gallbladder checked for gallstones by ultrasonography. Between May 1992 and June 1993, 1962 out of the 2235 (87.7%) subjects without gallstones at baseline agreed to a further ultrasound examination. At the first survey a standardized questionnaire was administered, inquiring about medical history, diet, cigarette smoking and other behavioural characteristics. Height and weight were also measured, and blood levels of glucose, cholesterol, HDL-cholesterol and triglycerides were determined by standard methods. The same variables were measured at the second survey. The diagnosis of gallstones was made with the same echograph by echographists working in the same department. Multiple logistic regression was used to determine which factors measured at the first survey were associated with the incident cases of gallstones. One hundred and four subjects had developed gallstones, an incidence of 9.7 per 1000 persons per year. Age, body mass index (BMI), weight change, a history of diabetes, constipation (shown by use of laxatives), cigarette smoking, years of schooling, consumption of fried foods and excessive oil, and pregnancy in females, were positively associated with the incidence of gallstones. Consumption of wine, coffee, fish and wholemeal bread was inversely associated. Sex, family history of cholelithiasis, use of oral contraceptives and serum lipids were not independent risk factors for gallstones. The results of this study confirm many gallstone-associated factors reported in previous cross-sectional and case-control studies, as well as in other cohort studies based on the clinical diagnosis of gallstones, such as BMI, ageing and wine consumption. Furthermore, use of laxatives, considered a proxy of constipation, appears to be another important independent risk factor for gallstones.