We test the empirical validity of the Capital Asset Pricing Model (CAPM) on the Zimbabwe Stock Exchange (ZSE) using cross-sectional stock returns on 31 stocks listed on the ZSE between March 2009 and February 2014. We conclude that, although the explanatory power of beta tends to fall rapidly for prediction horizons greater than 6 months, beta significantly explains average monthly stock returns on the ZSE. Tests to validate the CAPM reject its validity for the ZSE however, primarily due to liquidity and skewness anomalies. We nevertheless fail to detect any size effects. There is encouraging evidence to suggest that the CAPM performs reasonably well in predicting average monthly returns over prediction horizons of between 3 and 6 months. We recommend that investors and analysts must exercise extreme caution in applying the CAPM. Furthermore, we discourage strategies based on the existence of a size premium on the ZSE. Instead, investors may consider neglected and negatively skewed stocks, albeit over appropriate horizons. Further research on other African stock markets will help verify if the optimal performance range of the CAPM is indeed 3 to 6 months. Development of standard continental proxy market portfolios will also improve the estimation of betas and enhance results of cross-country tests of the CAPM.