The prognosis is poor for patients with pulmonary artery stenosis-associated pulmonary hypertension (PAS-PH). Identifying predictors of prognosis in PAS-PH is crucial to preventing premature death, which has rarely been investigated. We aimed to explore the cardiopulmonary exercise testing (CPET) parameters to predict the prognosis of these patients. We prospectively included all patients with PAS-PH who underwent CPET between September 2014 and June 2021 in Fuwai Hospital (ClinicalTrials.gov ID: NCT02061787). The primary outcome was clinical worsening, including death, rehospitalization for heart failure, or deterioration of PH. Seventy-two patients were included in this study. A median of 2-year follow-up revealed that 18 (25%) patients experienced clinical worsening. The 1-year, 3-year, and 5-year event-free survival rates were 92.5%, 81.7%, and 62.7%, respectively. Patients with clinical worsening demonstrated significantly worse baseline haemodynamics and poorer exercise capacity than their counterparts. Multivariable Cox regression identified that peak O2 pulse could independently predict clinical worsening [hazard ratio: 0.344, 95% confidence interval (CI) 0.188-0.631, P<0.001], outperforming other parameters. Peak O2 pulse correlated with PH severity. Incorporating peak O2 pulse into the simplified 2015 European Society of Cardiology/European Respiratory Society risk stratification improved the accuracy for predicting clinical worsening (pre vs. post area under the curve: 0.727 vs. 0.846, P<0.001; net reclassification index: 0.852, 95% CI 0.372-1.332, P<0.001; integrated discrimination index 0.133, 95% CI 0.031-0.235, P=0.011). The prognosis is poor for PAS-PH, and exercise intolerance and ventilation inefficiency are commonly observed. Peak O2 pulse independently predicted the prognosis of these patients. A low peak O2 pulse identified patients at high risk of clinical deterioration and served for risk stratification of PAS-PH.