OBJECTIVE The purpose of our study was to review and report the patient selection, techniques, and results of percutaneous drainage of pancreatic abscesses by retrospective review. METHODS Fifty-nine patients (46 men and 13 women) with a mean age of 44 years old had 80 pancreatic abscesses that were drained percutaneously under radiologic guidance (CT, n = 77; sonography, n = 2; and fluoroscopy, n = 1). Abscesses had a wide spectrum of causes, with alcoholic pancreatitis being most common, trauma second most common, and gallstones third. Ten patients had undergone surgery for pancreatic necrosis or abscess. Patients with pancreatic pseudocysts, necrosis, or acute fluid collections were excluded from this study. RESULTS Of the 59 patients, 51 (86%) were cured with percutaneous drainage and antibiotic therapy. Of the patients who were not cured with percutaneous drainage, seven required surgery and one underwent repeat percutaneous drainage. In the 59 patients, complications included non-life-threatening bleeding in three patients. Ten of 59 patients (17%) had fistulas that spontaneously formed into the gastrointestinal tract. The duration of catheterization ranged from 4 to 119 days, with a mean duration of 33 days. The rate of mortality at 30 days after completion of percutaneous drainage was 8% (5 of 59). CONCLUSIONS Percutaneous drainage was an effective therapy for this defined group of patients with pancreatic abscesses. Factors leading to the relatively high success rate described in this study likely included selection of patients; catheters of adequate size, number, and location; careful follow-up with appropriate catheter manipulations; and an integrated, cooperative approach whereby surgeons were willing to permit drainage to effect its benefits, rather than operating prematurely.