OBJECTIVES: Prior studies of direct costs of back pain have primarily used Workman's Compensation claims, resulting in varying cost estimates and limiting generalizibility of findings to specific employee groups. The objectives of this study were to determine the direct costs of back pain among the general population and to stratify those costs by type of medical care. METHODS: Retrospective analysis was conducted of the 1996 Medical Expenditure Panel Survey. The survey files included data collected from a nationally representative sample of 22,601 respondents and from respondents' medical care and health insurance providers. Data included medical conditions, use and payments for medical care, and health insurance and employment information. Back pain patients were identified using ICD-9-CM codes determined by an expert panel of physicians and coders as indicative of back pain. Direct costs were calculated using patient and third-party payments for back pain related medical events by category of back pain and type of medical care. Sample estimates were projected to the population and 95 percent confidence limits were calculated using the Taylor expansion method. RESULTS: The direct costs of aggregate back pain were $14,701,417,650, work-related back pain were $2,643,650,647, and missed workdays back pain were $5,396,355,220. With regard to aggregate back pain, higher proportions of direct costs were incurred for office-based medical provider visits at $5,768,302,826 (mean=$68; 95% C.L.=$61 to $75), inpatient visits at $4,658,655,867 (mean=$10,016; 95% C.L.=$9,463 to $10,570), and outpatient visits at $2,050,207,343 (mean=$274; (95% C.L.=$204 to $345). Lower proportions of direct costs were incurred for prescriptions, home health visits, and emergency room visits. CONCLUSIONS: Direct costs associated with back pain are sizable. As direct costs of occupationally related back pain were substantially lower than total direct costs of $14.7 billion, Workman's Compensation claims may underestimate the true impact of back pain on medical care utilization and expenditure.