In this paper, we analyze the spatial distribution of US employment and earnings against an urban wage-efficiency background, where leisure and effort at work are complementary. Using data from the American Time Use Survey (ATUS) for the period 2003-2014, we analyze the spatial distribution of employment across metropolitan areas. We also empirically study the relationship between individual earnings and commuting and leisure. Our empirical results show that employment is mostly concentrated in metropolitan cores, and that earnings increase with "expected" commuting time, which gives empirical support to our urban wage-efficiency theory. Furthermore, we use Geographical Information System models to show that there is no common pattern of commuting and the employees-to-unemployed rate, although we find higher wages in comparatively crowded states, where average commuting times are also higher.