A study of the labor exchange performance of the Employment Service showed that even after adjusting for differences in economic and demographic conditions, local offices and states varied greatly in their ability to place applicants in jobs. The performance measures examined were placement rate, permanent placement ratio (defined as the percentage of placements in jobs expected to last over 150 days), and placement wage ratio (defined as the average placement wage as a percentage of average community wage). Local offices that were stronger performers tended to be concentrated in certain states, indicating that the policies and practices of individual states may contribute to the variations in performance. The concentrations were not related to the level of resources extended per applicant. States with higher concentrations of above average performing offices included Florida, Georgia, Massachusetts, New Hampshire, North Carolina, and South Carolina, all of whom had twice the rate of offices with above average performance on all three measures as compared with the national average. Michigan, New Jersey, Rhode Island, and West Virginia all had more than double the national percentage of offices with below average performance for all three measures examined. Although Employment Service workload stayed about the same from 1980 to 1987, fewer offices provide one-on-one assistance during intake or providing counseling or testing services. (Includes 7 tables, 18 figures, and 13 references.) (CML)