Objective This article aimed to validate and compare the prognostic performance of generic scores (Pulmonary Embolism Severity Index [PESI] and Hestia) and cancer-specific pulmonary embolism (PE)/venous thromboembolism (VTE) scales (Registro Informatizado de la Enfermedad TromboEmbolica [RIETE], POMPE-C, and modified Ottawa) in PE patients with active cancer. Methods A retrospective study was conducted among 460 patients with PE and active cancer. The primary outcome was 30-day overall mortality. Secondary outcomes were 30-day PE-related death and overall adverse outcomes. The prognostic accuracy of clinical scores was determined using receiver operating characteristic (ROC) curve analysis. Results Within 30 days, 18.0% of patients died, 2.0% suffered major bleeding, and 0.2% presented recurrence of VTE. All scales showed a high area under the ROC curve (AUC) for predicting 30-day overall mortality except modified Ottawa (0.74 [0.70-0.78] for PESI, Hestia, and RIETE; 0.78 (0.74-0.81) for POMPE-C; 0.64 (0.59-0.68) for modified Ottawa]. PESI divided the least patients (9.1%) into low risk, followed by modified Ottawa (17.0%). Hestia stratified the most patients (65.4%) as low risk. But overall mortality of low-risk patients based on these three scales is high (>5%). RIETE and POMPE-C both classified 30.9% of patients as low risk, and low-risk patients stratified by these two scales presented a low overall mortality (1.4 and 3.5%). Similar predictive performance was found for 30-day PE-related death and overall adverse outcomes in these scores. Conclusion Cancer-specific PE prognostic scores (RIETE and POMPE-C) performed better than generic scales (PESI and Hestia) and a cancer-specific VTE prognostic scale (modified Ottawa) in identifying low-risk PE patients with active cancer who may be suitable for outpatient treatment.