OBJECTIVES: The study objectives were to determine the prevalence and correlates of potentially inappropriate prescribing among the ambulatory elderly using 2003 Beers criteria. METHODS: Retrospective analysis was conducted of the 2001 public use files of the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey. A total of 7243 visits by individuals over 64 years old with at least one prescription were analyzed. The 2003 Beers criteria were used to define potentially inappropriate medications irrespective of disease, dose, and duration. Multivariate logistic regression using generalized estimating equations was performed to examine associations between age, gender, race, payment source, reason for visit, referral status, sharing of care by other physicians, number of medications, ambulatory setting type, metropolitan location of practice, and geographic region with potentially inappropriate prescribing. Data was analyzed using SAS 8.2 and SUDAAN 8.02. An alpha of 0.05 was required for significance. RESULTS: Potentially inappropriate medications were prescribed in 17.8 million office-based ambulatory visits and 930,211 hospital-based ambulatory visits, 11.9% of all ambulatory visits made by the elderly. The most common drug classes were narcotic analgesics, antihistamines, and antiarrhythmic agents. Among all variables examined, location of practice, referral status, and number of medications were associated with receipt of a potentially inappropriate medication. After adjusting for other risk factors, visits made in metropolitan areas or by referred patients were more than twice as likely to involve a potentially inappropriate medication. In addition, compared with patients taking one medication, those taking two (Odds Ratio (OR) = 2.64, 95% CI = 1.41-4.95), three (OR = 6.85, 95% CI = 3.15-14.88), or four or more medications (OR = 7.43, 95% CI = 4.36-12.67) were more likely to receive a potentially inappropriate medication. CONCLUSIONS: Potentially inappropriate prescriptions were prevalent in nearly 12% of ambulatory visits made by the elderly in 2001. More prospective efforts to improve prescribing practices and prevent drug-related problems among the elderly are needed.