Different methods have been used to define an effective and tolerable degree of pulmonary artery banding (PAB) intraoperatively. As the hemodynamic response determines the results, a reduction of risk could be expected by evaluating several selected parameters and adjusting the band accordingly. Between 1981 and 1986, PAB was performed in 12 infants, three with complex cardiac lesions and nine with complete atrioventricular canal defects. Pressures in the distal pulmonary artery, in a systemic artery or the aorta, and in a central vein were monitored, and oxygen saturation measurements in both great arteries were taken before, during, and after the procedure. Adequate, complicated, intolerable, and contraindicated constriction were recognized and shown on characteristic pressure traces and oxygen saturation shifts. The measurements provided an indirect but reliable indication of the change in intracardiac shunting, and predicted whether the banding would be hemodynamically tolerated. Postoperatively, ten of the infants improved significantly, especially the seven infants previously in a low cardiac output state. One infant died on the day of surgery from acute right heart failure (8% hospital mortality). The results suggest that hemodynamically controlled PAB is an effective palliative treatment with a lower risk than had been previously reported.