The prevalence and cost of type-2 diabetes is significant in elderly patients. Improved glycemic control may be associated with better health outcomes and lower cost. OBJECTIVE: To analyze health care resource use and estimate cost of care over a two-year period in elderly patients (>60 years) with type-2 diabetes receiving treatment with rosiglitazone (RSG) plus sub-maximal sulfonylurea (SU) combination therapy (n=115) or progressive uptitration of the SU, glipizide (GLIP), (n=110) in the Rosiglitazone Early vs. Sulfonylurea Titration (RESULT) clinical trial. METHODS: Treatment was individualized, targeting ADA defined goals, as appropriate, with uptitration required for FPG>180mg/dL to a max of glipizide 20mg bid and RSG 4mg bid. Patient self-reported hospitalizations, emergency room (ER) visits, and physician visits were prospectively collected for the duration of the trial. Health care utilization rates were reported and analyzed as rates per 1000 patient-days using Poisson regression models. National average unit costs were applied to estimate total direct medical cost, where appropriate costs were adjusted for the duration of therapy and expressed as cost per patient per month (PPPM). RESULTS: By the end of two years, disease progression (time to reach confirmed FPG=d 180 mg/dl) was observed in only two patients (1.7%) randomized to RSG + GLIP, compared to 27 patients (24.3%) taking GLIP alone (p<0.0001). In comparison with patients in the GLIP group, patients in the RSG+GLIP group had significantly fewer ER visits (p=0.0006) and hospitalizations (p=0.0263). There were no statistically significant differences in unscheduled physician office visits between the two treatment groups. Average PPPM costs were significantly lower for the RSG+GLIP group ($480) compared to the GLIP monotherapy group ($644) (p<0.05). CONCLUSION: The addition of RSG to SU therapy was associated with a decreased use of medical resources, in particular hospitalizations and ER visits, and resulted in significant cost savings.