OBJECTIVES: In The Collaborative Atorvastatin Diabetes Study (CARDS), atorvastatin 10mg reduced CHD (36%) and stroke (48%) in primary CVD prevention patients in the UK with type 2 diabetes compared to placebo. However the transferability of the clinical and economic benefits for treatment algorithms and populations beyond the UK are not known. Despite the fact that relative risk reductions are similar between different populations, baseline risk varies according to population. This study outlines an approach to estimate baseline risk in a non-study diabetes population and calculate cost-effectiveness therein using the trial treatment effect. METHODS: A Markov model was developed using the UKPDS risk function to model the incidence of CHD and stroke assuming “no treatment”. The model allows re-calibration of risk for non-study diabetes populations by substituting the UKPDS population, individual, and mean risk factors with those of the population of interest. The treatment effect is determined by multiplying estimated incidence rates by the hazard ratio observed for CHD and stroke in CARDS. Competing hazards are used to estimate non-CV mortality. Model endpoints include event rates, estimated costs, survival and QALYs gained associated with atorvastatin 10mg treatment. RESULTS Validity was tested by substituting the CARDS trial population individual and mean risk factors with those from the UKPDS risk function. The re-calibrated UKPDS risk function resulted in four-year “no treatment” CHD predicted risk of 6.70% vs. actual risk of 6.52% (97.4% of predicted value). The corresponding stroke incidences were 2.37% and 2.48% (104.8% of predicted value). CONCLUSION: This approach improves the external validity by using the annual adjustment of baseline risk and maintains the internal validity by using observed clinical trial treatment effects. The re-calibrated model accurately predicted the incidence rates observed in the CARDS trial and enhances generalizability of clinical results to a non-study, non-UK population of diabetes patients.