This paper examines the robustness of explanatory variables in cross-country economic growth regressions using a novel approach called Bayesian Averaging of Classical Estimates (BACE). BACE constructs estimates as a weighted average of Ordinary Least Squares (OLS) estimates for every possible combination of included variables, with weights justified on Bayesian grounds similar to the Schwarz criterion. The authors find that 11 out of 32 explanatory variables are robustly partially correlated with long-term growth, and another five variables are marginally related. The strongest evidence is for the initial level of real GDP per capita. The paper also discusses the statistical theory, implementation issues, and empirical results, concluding with a detailed analysis of the variables strongly and robustly related to growth.This paper examines the robustness of explanatory variables in cross-country economic growth regressions using a novel approach called Bayesian Averaging of Classical Estimates (BACE). BACE constructs estimates as a weighted average of Ordinary Least Squares (OLS) estimates for every possible combination of included variables, with weights justified on Bayesian grounds similar to the Schwarz criterion. The authors find that 11 out of 32 explanatory variables are robustly partially correlated with long-term growth, and another five variables are marginally related. The strongest evidence is for the initial level of real GDP per capita. The paper also discusses the statistical theory, implementation issues, and empirical results, concluding with a detailed analysis of the variables strongly and robustly related to growth.