The objective of this study was to evaluate the outcomes of a diabetes disease management initiative among TennCare's Medicaid Population. A quasi-experimental group design was conducted using a control group and a diabetes disease management intervention group. Primary outcomes measures were rates for three key recommended tests (ie, microalbuminuria, lipids, and hemoglobin A1c). Secondary performance measures --patient satisfaction and program evaluation issues -- also were assessed. The study was performed among TennCare beneficiaries with diabetes mellitus. It utilized a quasi-experimental nonequivalent control group design, with 993 intervention participants in Knoxville and 1167 control group members in Chattanooga. Variables analyzed included testing rates for hemoglobin A1c, lipids, microalbuminuria, and demographics. A logistic regression model using baseline covariates was constructed to analyze the differences between the intervention and the control groups. Intracluster correlations were accounted for by generalized estimating equations. Statistical process control detected process changes in testing rates over time. There were meaningful changes in the rate of ordering recommended tests. The odds of an individual in the intervention group having at least one microalbuminuria test were 196% more (confidence interval [CI] = 1.50, 5.82; p = 0.002); the odds of having at least one lipid profile were 43% more (CI = 1.01, 2.02; p = 0.042); and the odds of having two or more hemoglobin A1c tests were 39% more (CI = 0.87, 2.23; p = 0.165) than the odds of an individual in the control group. The analysis also showed a high rate of satisfaction among patients in the intervention group. The program was successful in meeting its stated goals of providing effective disease management for TennCare patients with diabetes.