100 patients with acute cholecystitis (AC) diagnosed by clinical, laboratory, and roentgenological examinations were randomly divided into 2 groups; early surgery (ES), operated within 7 days after the onset of acute symptoms, and delayed (DS), operated 2--3 months after the acute episode. Patients with elevated serum bilirubin and/or amylase were included in the trial. Two patients died during conservative treatment, and in 4 cases medical treatment was interrupted because of peritonitis, and in 3 cases because of increasing jaundice. Recurrence of AC was found in 24% of the patients in the DS group. There was no mortality in the ES group, but 2 patients died postoperatively in the DS group. Wound infection developed in 3 patients in the ES, and in 8 patients in the DS group. Retained stones remained in 3 cases of the DS and in 1 case of the ES group. The operative procedures were easier to perform in the ES group than in the DS, as estimated by the duration of operation. The operation time was 76.7 +/- 4.6 min (mean p S.E.) in the ES and 98.0 +/- 7.3 min in the DS group. There was a statistically significant difference between the 2 groups (p less than 0.01). The results suggest that early surgery in the treatment of acute cholecystitis is recommended. The complications of failed medical treatment can be avoided by early operation without added risk of mortality or complications.