The correct timing of surgery in cases of gallstone pancreatitis is debatable. To delineate more clearly the influence of the timing of surgery in the treatment of the disease, a prospective randomized clinical study of early surgery (less than 48 hours after admission) and delayed surgery (more than 48 hours after admission) was conducted in 165 patients. Ranson's prognostic signs of severity of disease were used to classify the patients into two risk groups: mild pancreatitis (three or fewer positive signs) and severe pancreatitis (more than three positive signs). In patients with three or fewer positive Ranson's signs, the time of surgery appeared to have little effect on the outcome, whereas in patients with more than three positive signs, early surgery resulted in a significant increase in rates of morbidity and mortality. Controlled randomization showed that in patients with gallstone pancreatitis, edematous or hemorrhagic necrotizing pancreatitis can develop, with or without impacted stones, early or late in the progression of the disease, during early or delayed surgery. These findings suggest that (1) although a gallstone initiates a bout of pancreatitis, it does not cause the progression of the disease; (2) the fate of the progression of pancreatitis is decided early by the amount of digestive enzymes being activated; (3) early removal of an impacted stone does not ameliorate the progression of pancreatitis; and (4) surgery should be performed during the initial hospital admission after the pancreatitis has subsided.