Although tricuspid valve z-scores have been used to predict outcome in pulmonary atresia with intact ventricular septum, they are statistically generated from local populations, and widespread generalization may not be appropriate. To determine if there are echocardiographic predictors of outcome that can be universally used, the records of all infants with this diagnosis since 1988 were reviewed for age, weight, type of surgery, and outcome. Preoperative and follow-up echocardiograms were reviewed for valve diameter and z-scores, and valve ratios were calculated. Thirty-six patients were divided into 2 groups: group 1 included 23 infants who had a successful biventricular repair; group 2 included the remaining 13 infants who did not have a successful repair. Preoperatively, both groups had similar ages, pulmonary, aortic, and mitral z-scores, and pulmonary/aortic ratios, but the patients in group 2 had significantly lower weight (3.5 +/- 0.6 vs 2.9 +/- 0.5 kg), tricuspid z-scores (-0.7 +/- 1.5 vs -2.3 +/- 1.2), and tricuspid/mitral ratios (0.8 +/- 0.2 vs 0.5 +/- 0.1). At similar follow-up, both groups of patients had similar weight, aortic and mitral z-scores, and pulmonary/aortic ratios, but group 2 infants had significantly lower pulmonary and tricuspid z-scores and tricuspid/mitral ratios. Compared with the preoperative echocardiograms, group 1 had significant increases only in pulmonary z-scores, and pulmonary/aortic and tricuspid/mitral ratios. Group 2 had no significant change in any echocardiographic variable. The tricuspid/mitral ratio was >0.5 in all group 1 infants, and in 6 of 13 group 2 infants (2 sepsis deaths, 4 palliations). Compared with a tricuspid valve z-score >-3, a tricuspid/mitral ratio >0.5 was a better predictor of biventricular repair. Thus, infants who have a successful biventricular repair have significantly greater preoperative weight, tricuspid valve z-scores, and tricuspid/mitral valve ratios. A tricuspid/mitral ratio >0.5 was the best predictor of a biventricular repair.