Using quarterly data to estimate state-level matching functions, we obtain point estimates that are slightly higher than are found with national gross flows data, likely because of inherent differences in the data sources. We also estimate matching functions separately by the source of the new hire, and show that the results are consistent with the assumptions of endogenous job search by the employed and a preference for employed applicants by firms. Thus, care must be taken in interpreting empirical matching functions, which are likely a reduced-form combination of a structural matching function and a job competition model.