The aim of the study was to examine the relationship between anastomotic leaks and the development of symptomatic anastomotic strictures after primary repair of oesophageal atresia (OA) with or without tracheo-oesophageal fistula (TOF) and the prognostic value of the anastomotic appearance on postoperative oesophagography. The records and X-ray films of 49 patients with OA with or without TOF admitted between January 1990 and June 1995 were reviewed retrospectively. Of these, 37 had a primary repair (34 immediate and 3 delayed) and a postoperative contrast swallow was done between day 5 and day 30 (median day 7). Radiological leaks were documented. In addition, the maximum transverse diameters of the upper pouch and anastomotic region and the length of the anastomotic narrow segment were measured; the ratio of upper-pouch diameter to that of the anastomosis was calculated (anastomotic index, AI). The need for dilatation of symptomatic anastomotic strictures and for repeat dilatation as a separate episode were documented. Of the 37 patients, 4 developed an anastomotic leak (11%), 3 radiological and 1 clinical. A symptomatic stricture occurred in 55% of patients. There was no correlation between a leak and later development of a symptomatic stricture, although the small number of patients with leaks precluded definite conclusions. No radiological leak progressed to a clinical leak. There was no correlation between the AI or the length of the anastomotic narrow segment and the need for dilatation, the need for repeat dilatation, or the number of dilatations. In this study, the presence of a leak, the AI, and the length of the narrow segment on early postoperative contrast swallow did not correlate with the later development of a symptomatic stricture. A radiological leak was of no clinical significance. Since major leaks are apparent clinically, an early routine contrast study did not influence management and is not necessary.