Cultural variables are incorporated into a baseline endogenous economic growth model. Cultural attitudes toward achievement and thrift have a positive effect on economic growth. Cultural attitudes concerning postmaterialism have a negative effect on economic growth. Ordinary least squares regression is used to test economic and cultural models of growth on a cross section of 25 countries. The encompassing principle is used to resolve competing theoretical specifications and to generate a final parsimonious model. A variant of Leamer's Extreme Bounds Analysis (EBA) is used to evaluate the sensitivity of parameter estimates. The conclusions are further supported by nonparametric methods including robust regression and bootstrap resampling. The data for the analysis are from the World Values Survey (1990) and from Levine and Renelt (1992). An empirical model that incorporates both cultural and economic variables is superior to an explanation emphasizing one set of these variables. The final model is robust to: (1) alterations in the conditioning set of variables; (2) elimination of influential cases; and (3) variations in estimation procedures.