Evaluation studies of worksite stress management training (SMT) are reviewed and methodological considerations are offered regarding (1) program orientation, (2) experimental design, (3) worker participants, (4) outcome measures, and (5) duration of training effects and worker compliance. Suggestions for additional research include the need to (1) employ additional comparison groups in order to detect training-specific effects, (2) evaluate SMT in blue-collar settings, (3) expand the scope of outcome measures to include employee behaviors, and (4) assess long term effects of SMT and factors associated with worker maintenance of learned skills. It is concluded that SMT has value as a prevention activity in work settings but its use as a treatment strategy for troubled workers is not supported conceptually or empirically. Companies and practitioners are encouraged to view SMT not as an isolated activity but as one component of occupational health and safety activities. A holistic approach which incorporates stress management into company health and safety philosophies is viewed as the optimal strategy.