OBJECTIVES: The observational, non-interventional EDGE study showed that vildagliptin is effective in patients with type 2 diabetes mellitus who have suboptimal glycemic control on metformin (MET) monotherapy in the real-world setting, confirming the results of previous randomized clinical trials (RCTs). Cost-effectiveness evaluations are typically based on RCT data, which offer high internal validity and are the gold standard in evaluating efficacy and short-term safety. Nevertheless, they lack external validity and generalizability and there is a growing trend towards the complementary use of real-world data. Consequently, we sought to perform a health economic evaluation of the EDGE study using an established diabetes outcomes model. METHODS: The IMS Core Diabetes Model (CDM), a recently validated diabetes model, was used to evaluate the costs and outcomes of MET+vildagliptin (M+V) compared to MET+sulfonylurea (M+S), based on data derived from the EDGE study worldwide. M+V was associated with HbA1c and BMI changes of -1.19% and 0.199 kg/m, respectively. Corresponding data for M+S were -0.99% and 0.707 kg/m, respectively. Published network meta-analysis data were used to populate the CDM with hypoglycemia rates. The model was run over a lifetime, with costs ($US) and benefits discounted at 3.0%. RESULTS: Predicted quality-adjusted life expectancy (QALE) was 11.14 and 11.07 in patients treated with M+V and M+S, respectively. Total direct costs were estimated at $US 90,788 and $US 85,692 for patients treated with M+V and M+S, respectively. Incremental differences between M+V and M+S were 0.07 for QALE and $US 5,096 for total costs, yielding an incremental cost-effectiveness ratio (ICER) of $US 72,800. CONCLUSIONS: In the real-world setting, compared to sulfonylureas, vildagliptin was estimated to be cost effective using US ICER thresholds. These data further highlight the potential role of real-world data in assessing health economic value.