To identify the prognostic factors correlating with the outcome of the perforated necrotizing enterocolitis (NEC), the charts of 20 premature infants with perforated NEC were reviewed. Eight patients had long-term survival and 12 died. Infants in the survival group had significantly lower incidence of starting to feed before perforation, patent ductus arteriosus, indomethacin use, acidosis, and shock. Sepsis was only found in the group of infants that died The survival group were diagnosed with NEC and intestinal perforation at an earlier age than the infants that died. The time from diagnosis of NEC to intestinal perforation was longer in the infants that died. The localization of necrosis from operativefindings were all confined to the ileum-cecum area in survival group. The infants in the survival group were likely to have a single perforation. In addition, according to the operative management, they were divided into three groups: Group I consisted of six patients who received peritoneal drainage only, group II consisted of three patients who received peritoneal drainage followed by laparotomy, and group In consisted of 11 patients who received primary laparotomy. Infants in groups I and II had significantly lower gestational ages and birth weights than group III. There were no significant differences in mortality rates among infants in groups I, II and III. We concluded that the factors including prior enteral feeding, patent ductus arteriosus, indomethacin use, acidosis, sepsis, shock, delayed onset of NEC and perforation, multiple perforation, and diffuse necrotic changes were significant prognostic factors of poor outcome. Peritoneal drainage was a resuscitative procedure in critical condition and laparotomy should mostly be considered as the final treatment in the infants with perforated NEC.