BackgroundThe RIETE‐VTE score was derived to risk‐stratify patients with cancer‐associated venous thromboembolism (CAT). ObjectivesTo externally validate the RIETE‐VTE score and to compare its prognostic performance with the modified Ottawa score. Patients/MethodsWe studied 178 elderly patients with CAT in a prospective multicenter cohort and assessed 30‐day all‐cause mortality, 90‐day overall complications (mortality, major bleeding, or venous thromboembolism [VTE] recurrence), and 6‐month VTE recurrence. Patients were stratified into RIETE‐VTE and modified Ottawa score risk classes (low, intermediate, high). We compared the discriminative power (area under the receiver operating characteristic [ROC] curve) to predict mortality, overall complications, and VTE recurrence. ResultsFifteen patients (8.4%) died within 30 days, 42 (23.6%) experienced an overall complication by day 90, and 6 (3.4%) had recurrent VTE within 6 months. The RIETE‐VTE and the modified Ottawa score classified similar proportions of patients as low risk (35.4% versus 31.5%; P = .37). No low‐risk patient died within 30 days. Low‐risk patients identified by the RIETE‐VTE and modified Ottawa score had similar rates of overall complications (7.9% versus 8.9%) and VTE recurrence (1.6% versus 1.8%). The modified Ottawa score and the RIETE‐VTE score had similar areas under the ROC curve for predicting all‐cause mortality (0.84 versus 0.75; P = .21), overall complications (0.74 versus 0.68; P = .26), and VTE recurrence (0.67 versus 0.64; P = .78). ConclusionsBoth the RIETE‐VTE and modified Ottawa score accurately identified elderly patients with CAT who are at low risk for short‐term mortality and who are potential candidates for outpatient care.