OBJECTIVES: To assess the impact of adding rosiglitazone (ROSI) treatment to metformin or sulphonylurea on the lipid profiles and daily medication costs (Med-costs) in type 2 diabetes patients (Type2-DM). METHODS: Type2-DM patients were identified based on the diagnosis or use of glucose lowering drugs during January 1,1999-June 30, 2004 using UK General-Practice-Research-Database (GPRD). ROSI patients were matched 1:1 to control patients using propensity scores and these patients were followed from the first date of the add-on therapy until the end of the 18-month follow-up period. A1C, cholesterol, and Med-costs of lipid and glucose lowering medications were compared before and after add-on therapy between the cohorts. Changes in the study variables were tested using t-test and analysis of covariance, and the mean and 95% confidence intervals for Med-costs were calculated by the bootstrap method. All Med-costs were based on 2007 UK pounds. RESULTS: The study matched 2669 pairs of patients, resulting in mean age of 59 and 44% female. None of the variables used for the propensity scores were significant after matching. After add-on therapy, a significant increase of the total cholesterol by 3.1% was found in the ROSI group, while a significant decrease of -6.4% was found in the control group (p<0.001). A1C was reduced by 0.77% in ROSI and 0.98% in control patients. A total of 734 patients added lipid lowering drug after index date in the ROSI group compared to only 600 in control group. The estimated increase of lipid lowering daily drug cost in the ROSI group after add-on therapy was greater than that of the control group, averaging £0.14 vs. £0.11 respectively. CONCLUSION: After addition of rosiglitazone to metformin or sulphonylurea, patients achieved reductions in A1C, but also significant increases in total cholesterol, which resulted in increased use of lipid lowering drugs with associated costs.