Classic economic theory predicts that markets will clear, leaving little or no gains from trade left on the table. Laboratory experiments have largely confirmed this, though the results of recent field experiments have been mixed, with some artefactual markets in developing countries performing relatively inefficiently. I create a realistic multi-round trading market in Uganda with market-experienced individuals to explore the efficiency of trading and test what individual traits predict market efficiency and bargaining success using a rich dataset on individual characteristics. In early rounds, market efficiency is low. By the final round, efficiency rates are closer to theory. I find that individual characteristics of the buyers and sellers strongly predict the level of efficiency within the individual rounds. Individual characteristics are also important for individual success and divide along bargaining power: for buyers, who by design have high market power, wealth and patience are positively and significantly correlated with rents; for sellers with low market power, education, anti-social behavior and aggression are positively and significantly associated with rents. The results of the bargaining game also correlate with wealth levels two years after the experiment, suggesting that market prowess predicts lifetime outcomes. The results add importance to the role of individual characteristics for individual and social efficiency outcomes.