This study's objective was to develop a methodology to apply pharmacoepidemiological research toward understanding and improving psychotropic drug use among children enrolled in Medicaid. Using Kansas Medicaid data for 1995-1996, we summarized drug claims, diagnoses, and demographics for children under 20 who received at least one psychotropic drug prescription over either year. The sequence of steps needed to assure a quality improvement role is discussed. Use of key personnel in less regulatory and more clinical data applications is critical. Illustrating this approach, we found disproportionate numbers of children receiving psychotropic drugs who were young boys and larger numbers of white children receiving psychotropic prescriptions relative to their Medicaid enrollment than either African-American or Hispanic children. Medicaid agencies can expand epidemiological capacity to understand service use among segments of the population they insure as part of an overall commitment to improving quality.