The existence of quasi-fixed costs of work may affect firms' desired employee hours and number of workers, which has important implications for the estimation of labor supply parameters. Firm-level data from the 1982 U.S. Employment Opportunity Pilot Project are used to estimate the importance to firms of employee quasi-fixed costs related to searching, hiring, training, and firing. Specifically, this paper examines how these costs affect number of workers and hours per worker, turnover, and vacancies, to the extent that the costs are determined by a firm's presumably exogenous industrial classification. Results show that higher costs are associated with lower turnover, fewer vacancies, and longer hours as predicted by a model of labor demand.