The delivery of reinforcing and corrective feedback following the safe or unsafe performance of workers in occupational settings has repeatedly demonstrated its promise as a method for reducing or preventing accidents or their precusors. Much, however, remains to be discovered about the optimal use of feedback. Among the relatively unexplored variables in this domain is to what extent involving the direct participation of workers as a source of delivery of feedback will influence safe performance on the job. Related findings from studies of workers' involvement in goal selection and other aspects of organizational policy and practice have yielded promising results. In the proposed study the main purpose, consequently, will be to examine the extent to which involving workers in the safety program by having them serve as the source of (primarily positive) feedback will enhance safe performance on the job. Specifically, we plan to assess safe job performance, train volunteer workers to supply performance feedback, and compare the results against the feedback practice currently in effect. It is hypothesized that such worker involvement will enhance or sustain high level of safe job performance, reduce accidents, and increase worker satisfaction with the program.