OBJECTIVES: To construct a lifetime model evaluating potential health benefits and costs applying to Scottish Type-2 diabetes mellitus patients initiating first-line oral monotherapy, for whom metformin is inappropriate because of contra-indications or intolerance. When lifestyle modification (diet and exercise) affords inadequate glycaemic control, these patients currently have no alternative to sulphonylurea (SU) therapy. The model compared novel agent pioglitazone (PIO) versus generic SU treatment. METHODS: A decision-analytic Markov model was constructed using published (UKPDS) cost data for diabetes management and co-morbidity treatment. Three prospective treatment pathways were explored: first-line PIO/second-line PIO+SU combination/third-line insulin; first-line SU/second-line PIO+SU combination/third-line insulin; and first-line SU/second-line insulin. The model incorporated efficacy evidence of glycaemic control under PIO and SU, measured as initial HbA1c improvements and the rate of disease progression in terms of HbA1c (the coefficient of failure). RESULTS: Patients treated with PIO achieved better HbA1c control and improved serum lipid profiles, which translated into fewer diabetic complications, better quality of life and improved overall survival. Additional drug costs of PIO over SU were partly offset by lower costs to treat and manage diabetes complications, and delayed insulin therapy. The estimated incremental cost per QALY gained of PIO was £2415 compared to SU (when followed by second-line PIO/SU and third-line insulin therapy). The incremental cost per QALY gained of PIO was £1514 compared to SU (when followed by second-line insulin therapy). CONCLUSIONS: Clinical trial evidence indicated superior glycaemic (HbA1c) control in patients treated with PIO, in comparison with those treated with SU. The model showed that PIO is a cost-effective intervention and thus a valuable addition to first-line treatment options for patients intolerant and/or contra-indicated to metformin. Importantly, initiating PIO as second-line combination treatment after first-line SU in this patient group was less efficient than providing PIO monotherapy in a first-line setting.