OBJECTIVES: Multiple medication use is common among individuals with Chronic Obstructive Pulmonary Disease (COPD). Specifically, use of antidepressants, inhaled corticosteroids (ICS), and statins may place individuals with COPD at high risk for new-onset diabetes. We examined the relationship between use of medications (antidepressants, ICS, and statins) and new-onset diabetes among Medicaid beneficiaries with COPD. METHODS: We used a retrospective longitudinal dynamic cohort design using multiple years (2005-2008) of Medicaid claims of beneficiaries with newly diagnosed COPD (n = 15,287).  We identified receipt of antidepressants, ICS, and statins using National Drug Codes (NDC) and new-onset diabetes using International Classification of Diseases-9^th-Clinical Modification (250.x2) codes. Chi-square tests of independence and multivariate logistic regressions were used to examine the relationship between medication use and new-onset diabetes. Instrumental variable approach was used to control for selection bias in statin use. We conducted analysis using SAS v. 9.3 and STATA v11. RESULTS: Among Medicaid beneficiaries with newly-diagnosed COPD, 6.3% had new-onset diabetes. After controlling for baseline characteristics, adults with ICS (AOR: 1.23; 95% CI: 1.07, 1.47) and statins (AOR: 1.48; 95% CI: 1.27, 1.72) had greater risk of new-onset diabetes compared to those without ICS and statins. However, analyses using combined medication categories revealed that only adults with statin use in combination with antidepressants and ICS or ICS alone were more likely to have new-onset diabetes. Instrumental variable adjusted regression which controlled for selection bias in statin use revealed no significant association between statin use and new-onset diabetes.  CONCLUSIONS: Our study results suggest that multiple medication use (antidepressants, ICS and statins) was not associated with new-onset diabetes. However, as individuals with COPD are at higher risk of developing new-onset diabetes due to shared risk factors, further research with longer follow-up and randomized design is require to evaluate the safety of these medications.