OBJECTIVES: The objective of the study is comparing the cost-effectiveness ratios, between vildagliptin 100mg, sitagliptin 100 mg, rosiglitazone 8mg and pioglitazone 30mg as add-on therapy in patients with failure to metformin. METHODS: A meta-analysis of published rosiglitazone, pioglitazone, sitagliptin and vildagliptin trials with failure to metformin were performed. A decision analytic model with a decision tree with Bayesian approach was developed. Cost-effectiveness analysis was made. Resources utilization included emergency room patients, outpatients and hospital inpatients services, drugs, etc. Data were obtained from hospital records. The information was validated by an expert panel. The unit costs were gotten from the Mexican Institute of Social Security (IMSS). The perspective was from IMSS. Analysis was conducted on a 12-month period and discounting rate was not used. An incremental cost-effectiveness ratio and incremental net benefits were obtained. One-way, two-way and probabilistic sensitivity analyses were performed and acceptability curves were constructed. RESULTS: The lower expected cost was with vildagliptin US$1,434, while higher expected cost was with pioglitazone US$5033, the lowest cost per succesfull unit was the one based on vildagliptin US$1304, while the highest was the one based on sitagliptin US $5136. ICER analyses show that vildagliptin was a dominant alternative over sitagliptin, pioglitazone and rosiglitazone. The incremental net benefits were higher for vildaglitin strategy in independent way to willingness to pay. Results were robust to sensitivity analyses. CONCLUSION: Vildagliptin 100mg was dominant therapy as add on therapy in patients with failure to metformin over sitagliptin, rosiglitazone y pioglitazone in DM 2 patients.