A variable will add to predictability if it captures time variation in risk premia or expectations of future dividend growth rates. In this paper, we propose a new predictive variable intended to capture risk premia, the demand for physical gold. We expect an increased demand for gold coins and bars to indicate a higher risk premium. Our expectation is based on the safe haven property of gold (Baur and McDermott. 2010; Coudert and Raymond-Fcingolcl, 2011), which implies that increased demand for gold is correlated with higher levels of risk or risk aversion. Despite the fact that the demand for gold seems to be an almost natural candidate as a proxy for the risk premium. there are no studies that use gold demand figures in attempts to predict stock market returns. The same applies to the literature on the safe haven effect of gold, which does not use actual gold demand but prices to test for its existence. One reason may be that the global aggregate demand for coins and bars is only available at a quarterly frequency from 2002. We use the quarterly coin and bar demand figures as well as two proxies that are available from the 19805 at a monthly frequency: sales of American Eagle gold coins and spreads between quoted prices of gold bars and the official gold fixing price. Since there exists a large variety of gold coins. Eagle gold coins can only serve as a proxy for the general coin demand, but the sales data are available over a longer period and at a higher frequency. The use of the price spread is motivated by the observation that during periods of increased physical gold demand, bar and coin dealers are often not able to fully satisfy the demand leading to a situation in which coins or bars are sold at a premium.‘ The empirical analysis confirms our hypothesis that estimates of the demand for gold enhance pre- dictions of expected equity returns, i.e. the larger the gold demand, the larger the expected equity return. The remainder of the paper is structured as follows. Section 2 describes the data, with a focus on the gold demand series. Section 3 presents the empirical results and Section 4 summarizes the findings and concludes.