Levine and Zervos (1998) presented cross-country econometric evidence showing that, in a sample of 47 countries, stock market liquidity contributed a significant positive influence on GDP growth between 1976-93. We show that the Levine-Zervos results are not robust to alternative specifications because of the incomplete manner in which they control for outliers in their data. We show that when one properly controls for outliers, stock market liquidity no longer exerts any statistically observable influence on GDP growth.