During a 44 month trial, 268 patients with wounds of the colon were entered into a prospective, randomized, nonblinded study. Consideration for primary closure demanded that: preoperative shock was never profound, blood loss was less than 20% of estimated normal volume, no more than two intra-abdominal organ systems had been injured, fecal contamination was minimal, operation was begun within eight hours, and wounds of colon and abdominal wall were never so destructive as to require resection. Once such criteria had been satisfied, colon wound management was dictated by last digit in the randomly assigned hospital number; odd indicated primary closure; even, exteriorization of the wound or primary closure with protection by a proximal vent. Results obtained in 139 determinant patients eligible for randomization revealed that primary closure (67 patients) had a lower infection rate of the incision (48% vs 57%, p > 0.05) and a still lower infection rate for the abdomen proper (15% vs 29%, p < 0.05) on comparison to the 72 patients with a randomized colostomy. Morbidity otherwise for the randomized colostomy was tenfold greater than if a primary closure had been performed. Average postoperative stay was six days longer (p < 0.01) if a colostomy had been created, exclusive of subsequent hospitalization for colostomy closure; while the total extra cost for management of the colon wound by colostomy was approximately $2,700.00. Although immediate mortalities were identical, one late death occurred following colostomy closure. These data not only confirm the safety of primary closure for colon wounds in selected cases, but also indicate that such should become the preferred method of treatment whenever specific criteria have been met.