This study analyzes changes in the regional subcultures of the United States using 2000 census and religious survey data. The results suggest a remarkable degree of continuity with those the authors identified in an earlier study. In addition, they demonstrate that a new multidimensional measure of state culture does a much better job in predicting social and political behavior than other frequently used indicators. Finally, they show how their new measure of state culture significantly reduces and often eliminates the problem of spatial autocorrelation in many state-level indicators that cannot be explained by differences in economic development and racial—ethnic diversity.