According to the EKH the anthropogenic pressure on environment is not expected to rise indefinitely as economy grows, but to decrease when the population wealth exceeds a certain level. The focus of this work is to investigate of the plausibility the Environmental Kuznets Hipotheses (EKH) for per capita CO2 emissions and per capita GDP, for OECD countries, on statistical grounds. From a statistical point of view if EKH was true we should observe an inverted-U relationship between per capita CO2 emissions and per capita GDP. Most of the literature faces this issue by adapting a quadratic or cubic linear or log-linear model to single countries or panels of countries. We use a different statistical model, suitable to show such kind of curvilinear association and at the same time to fit the observed data better than the usual ones. Moreover, we carry out the analysis within a hierarchical bayes setting with country specific parameters in order to deal with the issue of poolability of different countries.