Background: Antiparkinson drugs (APDs) are the mainstay in the management of Parkinson's disease (PD). However, the use of APDs has not been investigated in a nationally representative sample. Objectives: This study aimed to describe the prevalence and patterns of APD use and to identify determinants of APD use among elderly Medicare beneficiaries with PD. Methods: Data from the Medicare Current Beneficiary Survey and Medicare claims from 2000 to 2003 were used to identify beneficiaries ≥65 years of age with PD. The APDs included in the study were levodopa, dopamine agonists, monoamine oxidase-B inhibitors, amantadine, anticholinergic agents, and catechol- O-methyltransferase inhibitors. The annual prevalence of APD use was calculated; patterns of APD use were measured by drug class and type of drug regimens (monotherapy or combination therapy). Potential determinants of APD use included sociodemographic characteristics (age, sex, race, education, marital status, and annual income), self-reported health status, residential status (institution or community), prescription drug coverage, disability (functional, physical, and mental), and disease-specific factors (dementia, depression, and other comorbidities). Multivariate logistic regression analyses with generalized estimating equations were used to estimate odds ratios (ORs) of any APD use associated with the determinants. Results: The data revealed 571 beneficiaries with PD (30.1% aged ≥85 years, 55.0% female, 88.3% white), who were observed for a total of 924 person-years. The annual prevalence of APD use was 58.2% (538/924). Levodopa was the most frequently used agent, either as monotherapy or in combination with other APDs (85.5%; 460/538). Multivariate analyses found that patients 65 to 74 years of age (OR = 0.70; 95% CI, 0.49–0.99; P < 0.05) and those ≥85 years of age (OR = 0.57; 95% CI, 0.40–0.81; P < 0.01) were less likely to use APDs than were patients aged 75 to 84 years of age. Patients who had attained a high level of education (more than a high school diploma) were more likely to use APDs than were patients with a low level of education (a high school diploma or less) (OR = 1.51; 95% CI, 1.04–2.19; P < 0.05). In addition, institutionalization (OR = 1.78; 95% CI, 1.17–2.71; P < 0.01), prescription drug coverage (OR = 1.50; 95% CI, 1.15–1.94; P < 0.01), activities of daily living (ADLs) (OR = 1.47; 95% CI, 1.16–1.87; P < 0.01), and depression (OR = 1.25; 95% CI, 1.02–1.53; P < 0.05) were significantly associated with use of APDs. The probability of APD use was lower among those with dementia than among those without dementia (OR = 0.62; 95% CI, 0.48–0.80; P < 0.001). Conclusions: Almost half of the elderly Medicare beneficiaries with PD in this study did not use any APD between 2000 and 2003. Levodopa was the most frequently used APD, either as monotherapy or in combination with other APDs. The identified determinants of APD use (age, education, prescription drug coverage, ADLs, dementia, depression, and residing in an institution) may be helpful in developing interventions for this population.