The Health and Retirement Study (HRS) is a long-running panel survey with good measures of economic status, so it is the pre-eminent data set for studies about the economic status of the older population and economic preparation for retirement. However, the HRS expends considerably fewer resources on the measurement of out-of-pocket spending than other surveys such as the Medical Expenditure Panel Survey (MEPS) and the Medicare Current Beneficiary Survey (MCBS), which may result in its having relatively less accurate measurement of such spending. We compare the level and distribution of out-of-pocket spending in the HRS with similar measures in MEPS and MCBS in the population aged 65 or older. We find that the measures of out-of-pocket spending in the HRS are about 50% greater than those in MEPS at the mean, and very much greater at the upper points of the distribution. HRS and MCBS are in better agreement, although the HRS is higher at the mean and at the top of the distribution. The implication is that the level and risk of out-of-pocket spending on health care are exaggerated in HRS. Observation error in the HRS measurement relative to MEPS and MCBS is to be expected, but this does not explain the apparent bias. We conclude that researchers who use HRS 2004 or earlier should examine health care spending carefully, even on a case-by-case basis.