OBJECTIVE: The UK Prospective Diabetes Study (UKPDS) has proven the relevance of an intensive glucose control policy in type 2 diabetes. UKPDS 35 has provided the evidence of a significant association between diabetes-related complications and level of HbA1c. We used the UKPDS findings to assess the cost-effectiveness of a new fixed-dose combination of metformin and glibenclamide, compared to the "conventional" strategy of the UKPDS. METHODS: We developed a Markov model reflecting the management of two diabetic cohorts. The follow-up of a cohort of newly diagnosed 50-year-old patients, similar to the "conventional" group of the UKPDS, was simulated to follow HbA1c progression over a 10-year period. The second cohort, treated with metformin-glibenclamide, had the same demographic and clinical characteristics at baseline. The reduction rate of HbA1c under metformin-glibenclamide was extracted from a 20-week randomized double-blind trial. HbA1c level in the metformin-glibenclamide cohort was assumed to progressively converge with that in the "conventional" group. The occurrence of complications was modeled through risk functions linking HbA1c levels to a conditional probability using UKPDS 23 and 35 results. Results were computed in a French context using a payer perspective. Only medical costs were considered. A sensitivity analysis was performed on the reduction rate of HbA1c under metformin-glibenclamide between 0.8 (best case) and 1.5 (worst case). RESULTS: Cumulative medical costs amounted to EUR 7,240 in the "conventional" group versus EUR 7,759 in the metformin-glibenclamide group. A 6.1% decrease in the mean number of events per patient was obtained. The additional cost per life year saved was EUR 13,142 [9,924 (best case), 17,912 (worst case)] and the additional cost per complication-free year was EUR 5,736 [4,312 (worst case), 7,842 (best case)]. CONCLUSION: These results suggest that metformin-glibenclamide is cost-effective in the treatment of type 2 diabetes when compared with "conventional" therapy.