Objective: To examine differences in medication adherence to adjuvant hormonal therapy between white and black patients with breast cancer using different statistical techniques. The data source: Linked North Carolina Medicaid claims-Tumor Registry data (years 1999-2005). Methods: The study design was a retrospective cohort study of Medicaid enrollees with breast cancer newly starting adjuvant hormonal therapy (tamoxifen or aromatase inhibitor) between years January 2000 to December 2004. Medication adherence [measured as Medication Possession Ratio (MPR)] was assessed using patients' prescription refill records. The Medicaid claims and CCR data were merged using a probabilistic match algorithm. From the linked data, information on patients with ICD-9 codes for primary breast cancer was extracted. Results: Black patients had a 7% and 9% lower adherence rate as compared to white patients in the propensity score and regression method respectively. Stratification based on 80% cut-off point for the MPR showed that black patients were 21% less likely to be in the high adherence group. Conclusion: Results from the propensity score and regression analysis may agree so closely in this study because there was good overlap in the distribution of background characteristics for the white and black women with primary breast cancer enrolled in Medicaid.