We analyze whether efficiency wages operate in urban labor markets, within the framework proposed by Ross and Zenou, in which shirking at work and leisure are assumed to be substitutes. We use unique data from the American Time Use Survey (ATUS) that allow us to analyze the relationships between leisure, shirking, commuting, employment, and earnings. We confirm that shirking and leisure are substitutes, and present an estimate of this relationship, representing the only empirical test of the relationship between a worker's time endowment and shirking at work. Our findings point to the existence of efficiency wages in labor markets.