About 1000 WIN welfare clients were selected in five cities: Harlem, New Brunswick, Milwaukee, Wichita, and Tacoma, and half were randomly assigned to the Job Club program. Of the continuing clients, 87% of the Job Club sample obtained jobs vs 59% of the Control sample at the 12-month follow-up and 80 vs 48% at 6 months. The Job Club was more effective in each of the five cities, for men and women, for high school graduates or dropouts, for blacks, whites, and Spanish, for handicapped or nonhandicapped, veterans or nonveterans, the young and the older, and for those required to participate as well as those who volunteered. The jobs obtained by the Job Club clients were comparable to the Control clients' jobs in terms of mean salary, full-time status, and type of job, and were more likely to be enduring, nonsubsidized, and obtained by the job-seeker's own efforts. Job Club members obtained employment in a median of six sessions (mean of 11); 90% obtained jobs within 23 sessions. Follow-up questionnaire data indicated comparable job retention in the two samples but slightly greater advancement and job satisfaction for the Job Club clients. The method appears to assure employment to virtually all participating welfare clients.