The primary aim of this research is to examine the potential link between areas considered to have poor retail food access and spatial concentrations of diet-related health problems including type 2 diabetes, obesity and low birth weight in Leeds and Bradford, West Yorkshire, UK. The research involves three major steps : (1) basic spatial analysis of disease distribution over the study area, (2) introducing measures of food access and (3) identification of the food access-health relationship with potential policy analysis. This paper reports the framework for policy analysis using both spatial interaction and microsimulation models. The spatial interaction model is used to estimate areas of the study region which have poor access to food retailers. The microsimulation model incorporates these food access measures and health data as well as socioeconomic and demographic data from several surveys. The model will be first run to predict disease incidence using only demographic variables. Predicted results will be compared to known incidence rates. In areas where the model under-predicts the diseases, environmental factors (including retail food access) will be introduced as an attempt to improve the predictions. This will allow us to identify which variables influence residents' health. As the microsimulation model identifies the factors negatively influencing peoples' health, it provides a framework for further policy analysis. The model can be used to run ‘what-if’ scenarios to predict the health implications of changes to environmental variables that proved to impact residents' health. The results of these scenarios will provide valuable information to policy makers who are concerned with health inequalities and the health officials who must treat an ever-growing population of obese and type 2 diabetics.