OBJECTIVE: To assess the cost-effectiveness of pioglitazone (PIO) based combinations versus alternatives for patients with type 2 diabetes in Norway. METHODS: A published/validated model for type1 diabetes developed by IMIB was adapted to simulate long-term management, health outcomes, resource utilisation and treatment costs of patients with type2 diabetes. The model accounts for most complications in diabetes patients: nephropathy, retinopathy, acute myocardial infarction, angina pectoris, stroke and amputation. The analysis was done from third-party-payer perspective and costs figured relative to the year 2000. A 5% discount rate was applied and sensitivity analysis was done to test the results. RESULTS: Pioglitazone PIO 30 mg and metformin (MF) were associated withlonger life expectancy (16.10 years) than sulphonylureas(SU)/MF (15.24) or rosiglitazone (RSG)/MET (15.95). PIO 30 mg/SU and PIO 15 mg/SU are associated with the lowest number of serious complications per 100 patients treated. For every 95 patients treated with PIO 30 mg/MF rather than SU/MF or every 27 patients, respectively, for PIO,30 mg/SU rather MF/SU, one complication is avoided. Combinations of PIO 30 mg/SU, PIO 30 mg/MF and PIO 15 mg/SU are associated with lower mortality than the other treatment combinations available. Thus, for every 34 patients treated with PIO 30 mg/MF rather than SU/MF, one death will be avoided after 15 years of treatment. After discounting both costs and life years at 5%, the above incremental cost-per-life-year is 29,406 Norway Kroner (NOK) in comparison to SU/MF, but still PIO dominates the combination therapy with RSG 8 mg. The picture is similar in the case of PIO 30 mg/SU combination compared to MFSU, where in the undiscounted incremental cost per life year gained, PIO dominates, and the cost per life year gained is raised to 25,992 NOK after discounting at 5%. CONCLUSION: Pioglitazone -based treatment for patients with type2 diabetes improves survival and reduces complications and therefore represents a cost-effective use of health-care resources in Norway. Nonetheless, the results such as those of in the model must be confirmed by long-term observational studies.