The purpose of this research is to explain the variation in the utilization of drug therapy for the medical conditions of depression, high cholesterol, and hypertension between Hispanics, non-Hispanic blacks, and non-Hispanics whites using Oaxaca-type decomposition analysis based on logit estimates. We find that almost the entire share of the utilization differences in drug therapy between blacks and whites can be explained by the differences in the coefficients of observable characteristics, while the sources of the utilization difference between the whites and Hispanics are split between the differences in the observable characteristics and the coefficient estimates. This result implies that strategies to improve racial and ethnic disparities need to be tailored to each group by focusing on the specific factors that are attributed to causing the disparity.