Although the mortality associated with gastroschisis (GS) has fallen markedly over recent years, postoperative morbidity and the incidence of complications remain high. Many different factors may contribute to this morbidity; the aim of this study was to determine which factors contributed most. Measures of morbidity used were time to full oral feeding (FOF), time on parenteral nutrition (PN), age at discharge, and incidence of complications. Between 1969 and 1995, 44 neonates with GS were treated; there were 6 deaths. The average initial temperature of the patients who died was 34.6 degrees C compared with 36.0 degrees C for the rest of the group (P = 0.02). Staged repair and prematurity were associated with increased time to FOF, time on PN, and age at discharge (P < 0.001). When the corrected post-term age was used, the difference between preterm and term babies was no longer significant. Mode of delivery did not influence any measure of morbidity. Seventeen patients (46%) had complications related to PN administration and 18 (43%) developed complications related to their surgery. There were no significant differences in these measures of morbidity when comparing patients born in the first half of the study period with those born in the last half. Multivariate analysis revealed that time to FOF, time on PN, and age at discharge were all strongly independently associated with staged repair and with the presence of complications of PN (all F > 7.2 and P < 0.01). Mode of delivery, gestational age, admission temperature, the need for postoperative ventilation, and complications of surgery were not independently associated with any of the measures of morbidity examined. Our data suggest that term delivery and primary closure of the defect are likely to minimise the morbidity experienced by infants with GS.