SUMMARY We report on experience gained in the treatment of 158 cases of oesophageal atresia presenting during a period of 10 years. The factors influencing mortality were analysed. During the period studied there was a slight improvement in survival, and this was probably due mainly to improved preoperative preparation of those babies undergoing primary repair. At best, 'staging' was thought to have had little influence on the survival of poor risk cases. Midwives, obstetricians, paediatricians, surgeons, and general practitioners did not do all that they could have done to prevent morbidity and mortality in these babies. At least one-third of the 79 deaths could have been prevented, and several deaths were caused solely by lack of awareness of the possible complications and associated abnormalities.Although much progress has been made in the treatment of infants with oesophageal atresia since 1943 (Haight and Towsley), before which time practically all cases died, there is still an unacceptably high mortality rate for 'poor risk' cases. Formerly, technical errors were responsible for many deaths, but advances in operative techniques, preoperative management, and paediatric anaesthesia have succeeded in substantially reducing the overall mortality rate. Since it is unlikely that any new operation of major importance will be developed in the foreseeable future, it is necessary to look to other factors in order to improve survival in these infants, particularly the early nursing and paediatric care, the preoperative management and choice of operation by the surgeon, and the subsequent community and environmental care.