This paper analyzes the technical efficiency of Dutch nursing homes with respect to the use of labor inputs by means of Data Envelopment Analysis (DEA). In addition, the determinants of the efficiency scores are investigated using censored regression analysis. Special attention is paid to checking the robustness of the results to using different versions of DEA and to the econometric specification of the censored regression models. Fifty percent of the nursing homes are fully efficient, according to the theoretically preferred frontier with constant or decreasing returns to scale. There is some evidence of a trade-off between labor input efficiency and the quality of care.