OBJECTIVES: To compare the multivariable-adjusted logistic regression model with the propensity score-matched, propensity score-stratified, and propensity score-adjusted logistic regression models in estimating the effect of exposure to antidepressant agents in increasing the risk of type 2 diabetes mellitus. METHODS: A retrospective cohort study using the Texas Medicaid prescription claims database was conducted from January 1, 2002 to December 31, 2009. Patients aged 18-64 years with new prescriptions for antidepressants (exposed group) or benzodiazepines (unexposed group) and without diabetes at cohort entry were included in the study. Propensity scores, which predicted exposure to antidepressant agents, were used to create propensity score-matched, propensity score-stratified, and propensity score-adjusted logistic regression models.  RESULTS: A total of 44,715 patients formed the study sample. The risk estimates varied across different analytic methods. The propensity score-matched logistic regression model yielded the highest risk estimate (Relative Risk [RR] = 1.452; 95% Confidence Interval [CI]: 1.276 – 1.651), followed by the multivariable-adjusted logistic regression model (RR=1.319; 95% CI: 1.067 – 1.630) and the propensity score-stratified logistic regression model (RR=1.153; 95% CI: 1.033 – 1.287). The propensity score-adjusted regression model yielded the smallest risk estimate (RR=1.080; 95% CI: 0.968 – 1.205). CONCLUSIONS: Propensity score techniques using pharmacy claims data with a limited number of covariates yielded varying estimates of the treatment effect. The choice of the propensity score technique may influence the magnitude of the treatment effect estimate.