The paper provides Monte Carlo evidence on the performance of general-to-specific and specific-to-general selection of explanatory variables in linear (auto)regressions. In small samples the former is markedly inefficient in terms of ex-ante forecasting performance.