Our entire institutional experience with pulmonary atresia and intact ventricular septum (1965 through 1987) included 115 patients, 16 of whom died before surgical intervention. Fifty-six percent of surgical patients (n = 99) had angiographic evidence of right ventricle-coronary arterial connections. The early mortality in the surgical group was 27.2%, and the actuarial survival was 24.7% +/- 6% at 13 years postoperatively. Multivariate analysis indicated that the presence of ventriculocoronary connections (p = 0.037), a decreasing ratio between right ventricular and left ventricular pressure at the initial cardiac catheterization (p = 0.007), and lower weight at operation (p = 0.001) were incremental risk factors for postoperative death; the presence of Ebstein's anomaly was an additional risk factor in the overall experience (including patients not surgically treated) (p = 0.01). Nearly all long-term survivors underwent at least one reoperation, including right ventricular outflow tract reconstruction (n = 39) and thromboexclusion of the right ventricle (n = 9). The presence of severe stenosis or interruption of the proximal left anterior descending coronary artery system was a uniformly lethal risk factor for patients undergoing these procedures (p = 0.0003). We conclude that surgical procedures that successfully decompress the right ventricle will usually result in biventricular circulation in and long-term survival of patients with pulmonary atresia with intact ventricular septum not complicated by Ebstein's anomaly or extensive ventriculocoronary connections. Decompression or thromboexclusion of the right ventricle is contraindicated in patients with ventriculocoronary connections and a right ventricle-dependent coronary circulation.