Objective: Our aim was to test whether a unidirectional valve patch would provide benefit to early and long-term survival for patients with ventricular septal defect and severe pulmonary artery hypertension. Methods: Eight hundred seventy-six cases of ventricular septal defect with severe pulmonary artery hypertension were closed with or without a unidirectional valve patch and were classified as the unidirectional valve patch (UVP) group (n = 195) and nonvalve patch (NVP) group (n = 681), respectively. Propensity scores of inclusion into the UVP group were used to match 138 pairs between the 2 groups. Kaplan-Meier survival curves were constructed to compare early and long-term survival. Results: For the 138 propensity-matched pairs, there were 7 and 9 early deaths (in-hospital deaths) in the UVP and NVP groups, respectively. The difference in early mortality between the 2 groups did not reach statistical significance (χ² = 0.265, P = .6064). With a mean of 9.2 ± 4.92 years' and 2511 patient-years' follow-up, there were 6 late deaths in the UVP group and 7 late deaths in the NVP group. The difference in actuarial survival at 5, 10, 15, and 18 years between the 2 groups was not significant (log-rank test, χ² = 0.565, P = .331). The difference in the late mortality between the groups with or without a patent patch at the time of discharge did not reach statistical significance (χ² = 1.140, P = .2856). There was no difference between the 2 groups in the 6-minute walk distance assessed at the last follow-up (525.9 ± 88.0 meters for the UVP group and 536.5 ± 95.8 meters for the NVP group, F = 1.550, P = .214). Conclusion: A unidirectional valve patch provides no benefits to early and long-term survival when it is used to deal with ventricular septal defect and severe pulmonary artery hypertension. © 2010 The American Association for Thoracic Surgery.