OBJECTIVES: To investigate the survival impact of skeletal-related events (SRE) as a cluster of events and by type of SRE, including pathological fracture (PF), spinal cord compression (SCC), and bone surgery (BS), using three claims-based approaches. METHODS: We analyzed prostate cancer (PCa)-specific and all-cause mortality among elderly SEER-Medicare men diagnosed with metastatic PCa between 2000 and 2007. PF, SCC, and BS were identified from Medicare claims and used to define SRE using a base case and alternative approaches (AltA). The approaches included a base case: SRE follows claims with a bone metastasis (BM) ICD-9 diagnosis code; AltA-1: SRE includes BM diagnosis code on the SRE claim; AltA-2: SRE is not anchored to BM. A Cox model provided covariate-adjusted hazard ratios (HR) in the full sample (FS) and in a propensity-score matched sample (PSMS). RESULTS: Application of inclusion/exclusion criteria resulted in 7,062 patients in FS (1,776 in PSMS).  PCa-specific mortality was 54% during median (mean) follow-up of 609 (837) days. Using the base case approach, the prevalence of any SRE, SCC, PF, and BS were: 12.6%, 7.4%, 4.3% and 1.9%. SRE prevalence was 9.7% for AltA-1 and 17.1% for AltA-2. The HR (95% CI) associated with any SRE was 1.27 (1.16–1.39) for the base case approach, 1.31 (1.18–1.45) for AltA-1, and 1.07 (0.98–1.16) for AltA-2.  Results were qualitatively similar for all-cause mortality. The HRs for SCC and PF were statistically significantly associated with PCa-specific mortality across approaches.  The results for BS were sensitive to the SRE definition.  Results for SCC and BS were unchanged from the base case approach when using the PSMS. CONCLUSIONS: The prevalence of SREs and their association with survival in the elderly depends on the definition of SREs. A validated claims-based measure of SRE is warranted to better understand the burden of SRE among PCa patients.