Prescription drug coverage issues have generated considerable attention within health care and political agendas. To allocate scarce resources efficiently, it is important to identify the characteristics of individuals with the highest level of out-of-pocket expenditures for prescription medications. OBJECTIVE: To identify factors explaining the variance associated with individual out-of-pocket prescription drug costs. METHODS: We analyzed data from the household component and event files of the 1996 Medical Expenditure Panel Survey (MEPS), with brand-generic status information obtained from MULTUM(r). Of the 22,601 individuals, 14,015 (62%) incurred prescription expenditures. Out-of-pocket prescription costs were regressed on demographic factors, socioeconomic factors, perceived health status, type of health insurance coverage, perceived access to care, prescription drug characteristics, and health care cost data. Person level weights were included in the regression model. Variance estimation was performed to adjust for the complex survey design employed by MEPS. RESULTS: The mean out-of-pocket prescription drug cost was $166 (se=4.61). Minorities had lower out-of-pocket costs ($117, se=7.87) compared to Caucasians ($175, se=5.05), whereas uninsured consumers had greater out-of-pocket costs ($181, se=20.29) compared to consumers with private insurance ($156, se=4.76). Significant (alpha=0.05) predictors with positive coefficients included: age, poverty status, total number of prescriptions, and total out-of-pocket non-prescription health care costs. Significant predictors with negative coefficients included: education, race/ethnicity, covered prescription costs and percent of prescription drug expenditures for generic drugs. The overall model was significant (p<0.001) with an R2=0.44. CONCLUSIONS: Out-of-pocket prescription drug costs were associated with several sociodemographic and health care cost variables. However, no significant associations were found between health status and perceived access to care variables. These findings suggest that economic factors are important when assessing out-of-pocket expenditures but other variables such as prescription drug characteristics should not be omitted.