The long-term impact of persistent pulmonary vascular obstruction after pulmonary embolism (PE) remains unknown. Based on ventilation-perfusion lung scan performed at discharge and 3 months after a first PE, we aimed to investigate the prognostic value on 5-year adverse events of (1) residual pulmonary vascular obstruction (RPVO) at discharge (DIS-RPVO), (2) RPVO at 3 months (3M-RPVO), and (3) relative change in RPVO between the 2 scans (RC-RPVO). We performed a prospective, multicenter cohort study from January 2007 to December 2009 including patients who survived at least 3 months after a PE. RC-RPVO was defined as (DIS-RPVO - 3M-RPVO)/DIS-RPVO. The primary end point was a combined end point at 5 years, composed of all-cause death, recurrent venous thromboembolism, chronic thromboembolic pulmonary hypertension, heart failure, and rehospitalization for cardiac causes. Receiver-operating characteristic curves were computed to define thresholds of DIS-RPVO, 3M-RPVO, and RC-RPVO predictive of the primary combined end point at 5 years. Overall, 241 patients were included (high-risk PE: 11.2%, intermediate-risk PE: 51.8%, low-risk PE: 37%). Mean DIS-RPVO was 27.9 ± 15.1%, mean 3M-RPVO was 10.3 ± 10.8%, and mean RC-RPVO was 61.7 ± 33.4%. At 5 years, 112 patients (46.5%) experienced the combined end point. Both 3M-RPVO ≥15% and RC-RPVO ≤37.5% were independently related to the occurrence of the combined end point at 5 years (p = 0.01 and p = 0.02, respectively). DIS-RPVO did not predict long-term adverse events. In conclusion, RC-RPVO ≤37.5% and 3M-RPVO ≥15% were independently related to the occurrence of adverse events 5 years after a first PE.