This paper develops a model of Temporary Help Agency (THA) sorting when workers are heterogeneous across skill and impatience. The presence of THAs is shown as a Pareto improvement in two ways. First more workers are employed in equilibrium with a THA. Second, a larger fraction of these workers are sorted into jobs which match their skills in the presence of THAs. This improved sorting technology makes workers better off in terms of their labor income. Empirical analysis from NLSY79 data shows that female agency temp workers sorted by the THA experience increased earnings compared to other temp workers, while men do not. To the degree that female THA workers have more job market patience compared to the men in this data, model predictions are consistent with wage impacts. The wage heterogeneity within each worker skill group arises as temp workers are sorted differently based on impatience.