Theoretical considerations appear to support the conjecture that stock returns are positively related to growth in the long run. However, the empirical literature does not give unanimous support to the theory. Based on a stochastic general equilibrium model it is argued that the long-run relationship between stock returns and per capita income growth is ambiguous and depends on output volatility. Using a century of data for 20 Organization for Economic Co-operation and Development (OECD) countries it is shown that the relationship between stock returns and growth is positive over the period 1916--1951, in which output volatility was persistent. Outside this period no relationship between stock returns and growth is found. These findings are consistent with the predictions of the theoretical model.