During the cross-sectional studies (February 1981 to July 1984) performed by the Group for Epidemiology and Prevention of Cholelithiasis (GREPCO) in Rome, Italy, 161 subjects were identified as having gallstones. Ten subjects did not participate in the prospective follow-up. At entry, 33 of the 151 remaining subjects were symptomatic, and 118 were asymptomatic. Data on incidence of biliary colics, complications, cholecystectomy, and death were collected at least every 2 years. In the initially asymptomatic group, the cumulative probability (% +/- SE) of developing biliary colic was 11.9 +/- 3.0 at 2 years, 16.5 +/- 3.5 at 4 years, and 25.8 +/- 4.6 at 10 years. None of the variables considered as possible modifiers of the natural history were found to be associated with an increased risk of developing biliary colic. The cumulative probability (% +/- SE) of developing complications after 10 years was 3.0 +/- 1.8 in the initially asymptomatic group and 6.5 +/- 4.4 in the symptomatic group (P = NS). Incidence of cholecystectomy was higher in the initially symptomatic than in the asymptomatic group (log-rank test = 2.27; P = .02). Fifteen (53.6%) of the 28 operated in the initially asymptomatic group were submitted to cholecystectomy, although specific symptoms did not occur. Twelve (10.2%) and 2 (6.1%) of the initially asymptomatic and symptomatic subjects died during the follow-up. One patient in the former group died at age 64 of a histologically proven adenocarcinoma of the gallbladder. In conclusion, this study demonstrates that the natural history of gallstones is less benign than is generally considered.