This study evaluates the role of primary peritoneal drainage (PPD) in the management of neonatal necrotizing enterocolitis (NEC). Of 169 patients with definite NEC, 92 (55%) underwent operation: primary laparotomy, 41 patients (45%); and PPD, 51 patients (55%). Seventeen (33%) of the PPD infants had subsequent laparotomy within seven days. Pneumoperitoneum was the indication for operation in 37% of the primary laparotomy and 67% of the PPD infants. Following PPD, 34 patients (67%) showed clinical improvement. Operative survivals were as follows: primary laparotomy, 83%; PPD, 53%. Infants who had PPD had a significantly lower mean birth weight, gestational age, preoperative pH and platelet count, and a significantly higher incidence of intraventricular hemorrhage and patent ductus arteriosus. For infants weighing less than 1,000 g at birth, the survival was similar following primary laparotomy (57%) and PPD (52%); this occurred in spite of the higher incidence of adverse risk factors in the PPD infants. For infants weighing greater than 1,000 g, the survival was 86% following primary laparotomy and 62% after PPD; in this group, all the early deaths following PPD occurred in critically ill infants who died within 48 hours of drainage. The late survival rates were as follows: primary laparotomy, 76%; PPD, 35%. More than half of the late deaths following PPD were not related to NEC, reflecting the difference between the two patient populations. Primary peritoneal drainage is a useful adjunct to resuscitation of the critically ill infant with complicated NEC, particularly prematures less than 1,000 g birth weight with intestinal perforation. Primary peritoneal drainage is not an alternative to laparotomy, which is recommended when an optimal clinical response has been achieved.