Objective To estimate the cost of defensive medicine among elderly Medicare patients. Data Sources We use a 2008 national physician survey linked to respondents’ elderly Medicare patients’ claims data. Study Design Using a sample of survey respondent/beneficiary dyads stratified by physician specialty, we estimated cross‐sectional regressions of annual costs on patient covariates and a medical malpractice fear index formed from five validated physician survey questions. Defensive medicine costs were calculated as the difference between observed patient costs and those under hypothetical alternative levels of malpractice concern, and then aggregated to estimate average defensive medicine costs per beneficiary. Data Collection Methods The physician survey was conducted by mail. Patient claims were linked to survey respondents and reweighted to approximate the elderly Medicare beneficiary population. Principal Findings Higher levels of the malpractice fear index were associated with higher patient spending. Based on the measured associations, we estimated that defensive medicine accounted for 8 to 20 percent of total costs under alternative scenarios. The highest estimate is associated with a counterfactual of no malpractice concerns, which is unlikely to be socially optimal as some extrinsic incentives to avoid medical errors are desirable. Among specialty groups, primary care physicians contributed the most to defensive medicine spending. Higher costs resulted mostly from more hospital admissions and greater postacute care. Conclusions Although results are based on measured associations between malpractice fears and spending, and may not reflect the true causal effects, they suggest defensive medicine likely contributes substantial additional costs to Medicare.