The Job Training Partnership Act (JTPA) was heralded as a major overhaul of the Comprehensive Employment and Training Act (CETA) program. Five years after passage of JTPA, the answer to the question of whether it has indeed managed to remedy CETA's shortcomings appears to be an unequivocal maybe. Although the JTPA has been praised by numerous business representatives and politically conservative public figures and government officials, a careful assessment of JTPA programs reveals that their performance falls far short of the claims made by administration officials and many program managers. Evidence suggests that JTPA's superiority over CETA may be illusory. For one thing, local administrators and training contractors select a more qualified clientele than CETA served. Moreover, by offering briefer and less intensive courses, JTPA does too little to improve the saleable labor market skills of its enrollees. JTPA's emphasis on outcomes also creates strong temptations to "doctor" program outcome data. Because of federal and state negligence, JTPA's dislocated worker projects have spent only two-thirds of their appropriated funds. Dislocated worker programs run under JTPA have tended to exclude the least educated and older workers (the ones most in need of help). In contrast, the Job Corps program seems relatively unchanged since the JTPA's passage, and efforts are now underway to replicate the highly effective Job Corps model of combining remedial and vocational education. Training programs for migrant farm workers and Native Americans have been particularly neglected under JTPA. Evidence suggests that, with reform, the program could improve considerably, even without altering its administrative structure. (This document contains extensive reference notes and an index.) (MN)