This article provides a new test of the predictive ability of aggregate financial ratios. Predictive regressions are subject to small-sample biases, but the correction in previous studies can substantially understate forecasting power. Dividend yield predicts aggregate market returns from 1946 - 2000, as well as in various subperiods. Book-to-market and the earnings-price ratio predict returns during the shorter 1963 - 2000 sample. The evidence remains strong despite the unusual price run-up in recent years.