OBJECTIVES: A number of diabetes models have recently been published. They are often used to assess the cost-effectiveness of new interventions and to generate health economic arguments for reimbursement submissions. The majority of these models do not account for rates of screening for important diabetes-related microvascular (eye, renal and foot) disease, nor do they consider the rates of use of important concomitant medications like ACE inhibitor/angiotensin-2-receptor inhibitors, statins, or aspirin. Our aim was to test the hypothesis that not accounting for these important factors may substantially influence projected long-term cost-effectiveness of new interventions. METHODS: A published and validated diabetes model was used to project the long-term cost-effectiveness of a hypothetical intervention that lowered HbA1c by 0.4%-points, and which cost an additional $500/patient/year, versus no intervention. Quality-adjusted life years (QALY) and lifetime direct medical costs were calculated for each treatment arm, taking into consideration: A) no screening for- and appropriate treatment of diabetes-related complications, or B) screening rates and concomitant medication use as seen in a typical type-2 diabetes population in the US. RESULTS: If screening rates and concomitant medication use were not considered, the hypothetical intervention was dominant to no intervention, with 0.214 QALYs gained (discounted 3% annually), and discounted lifetime direct cost savings of $165/patient. When screening rates and concomitant medication rates were accounted for, the intervention led to smaller improvements in QALYs, and increased costs, with incremental costs/QALY gained of $34,024. CONCLUSIONS: Health economic models of diabetes must account for the costs and clinical effects of screening for- and appropriate treatment of important diabetic microvascular complications, and the costs and effects of important concomitant medications. Failure to account for these factors may lead to inaccurate assessment of the cost-effectiveness of new interventions in type 2 diabetes patients.