OBJECTIVES: To describe use of estrogen replacement therapy (ERT) among Maryland Medicaid females and to identify factors associated with its use and discontinuation. METHODS: This study was conducted using Medicaid claims data to identify women, age 45 years or older, with at least two years of continuous Medicaid eligibility between October 1, 1995 and September 30, 1998. Drug use was based on prescription claims and clinical characteristics were based on medical, institutional, and prescription claims. RESULTS: 60,531 female recipients were eligible for the study with a mean age of 68 years (SD=12.9). The overall ERT prevalence was 14%. ERT use was the highest in the 45-54 age group (23.9%, P<0.001) and higher among white as compared to nonwhite women (15.7% vs. 12.9%, P<0.001). Unopposed estrogen products were dispensed to 70.5% of ERT users (average duration=142 days). The remainder used estrogen-progestin combinations and had shorter duration (average=85 days). ERT use was higher among women with osteoporosis, ischemic heart disease, hypertension, hyperlipidemia, or diabetes as compared to patients without these conditions (P<0.001). ERT use was low among breast cancer patients (P<0.001), whereas a diagnosis of endometrial cancer, or thromboembolic disease was unrelated to ERT use. Using proportional hazards models, discontinuation of ERT was associated with nonwhite race (P<0.001) and estrogen sensitive cancers (P<0.001). Conversely, having claims indicative of osteoporosis, ischemic heart disease, hypertension, or hyperlipidemia was unrelated to discontinuation of ERT. CONCLUSION: ERT use in this population was low compared to estimates for US women 50 or older based on NHANES III. Intervention should be considered to increase ERT use among women with cardiac risk factors and to improve long-term compliance.