Although the rational expectations hypothesis is widely applied in asset pricing models with differentially informed traders, the extent to which markets actually aggregate and transmit information is an open question. In this study we report the results from laboratory asset markets which were designed to examine this question. Initially we designed our markets with the intention of giving the RE model its best chance for success. After documenting evidence in favor of the RE model we examined which features of our environment are necessary or sufficient to achieve an RE equilibrium. Within the class of treatments we consider we find that trading experience and common knowledge of dividends are jointly sufficient to achieve an RE equilibrium, but that neither is a sufficient condition by itself. We also present some stylized facts about the convergence process leading to an RE equilibrium.