Poor diets and rising obesity rates dominate the current food, nutrition and health policy debate in many countries, including Spain. Despite the increasing obesity rate in Spain, there has been no known published research in Spain that has studied the economic factors affecting obesity prevalence. The main aim of our work is analysing the relevance of economic factors in obesity prevalence in Spain. This aim is especially relevant in shadow of the Economic crisis that hit Spain recently. Our methodological approach depending upon the estimation of a multinomial Logit Regression (MLR) Combined with a nonparametric model, the Multivariate Adaptive Regression Splines (MARS), to judge the role of different socioeconomic factors on the obesity prevalence. Despite the desirable advantages of using nonparametric models such as MARS, our paper is the first attempt to use this type of models to analyse the determinant factors of obesity prevalence. Our results suggest that Socio Economic factors seem to have a significant impact on obesity prevalence. MARS models outperform the traditional MLR and could be a helpful tool for understanding the nature of the relationship and moreover it could be helpful as pre estimate guides the estimation of its parametric counterpart.