A high divided sigmoid colostomy has been recommended for staged management of high anorectal malformation. We audited our cases of neonatal colostomy for high anorectal malformation to assess its effectiveness. A retrospective study was carried out of all surgical newborns admitted with high imperforate anus as the single diagnosis at our centre between December 1998 and December 2000. Morbidity and mortality were analysed after retrospective stratification into two groups (group A: birth weight >2.5 kg; group B: birth weight <2.5 kg). The chi square test was used to test the statistical significance in terms of outcome in the two groups. Overall mortality was 16%. Group A consisted of 34 babies: 30 with divided sigmoid colostomy and four with transverse loop colostomy. One baby with a divided sigmoid colostomy died from wound complications and septicaemia (mortality 2.9%). All four babies with transverse loop colostomy done under local anaesthesia survived, despite being sick on arrival. Group B consisted of 16 babies: 15 with sigmoid colostomy and one with transverse loop colostomy, with seven deaths (44%). None of the five babies with transverse loop colostomy done under local anaesthesia died, despite being sick on arrival, whereas all eight babies who died had undergone sigmoid colostomy under general anaesthesia. The difference in the outcomes of babies in groups A and B is highly significant ( p <.01). Sick, small (<2.5 kg) and septic babies arriving late to the unit do not appear to tolerate general anaesthesia and divided sigmoid colostomy well, despite that procedure's long-term advantages. Divided sigmoid colostomy has produced excellent results in babies >2.5 kg, but in the context of the developing world and limited critical care availability, transverse loop colostomy under local anaesthesia may save lives.