Field research on the relationships among environmental stressors, psychological reactions, and health outcomes has either been correlational or of a program evaluation nature. Neither of these alternatives is satisfactory for testing hypotheses concerning the causal status of variables in process models of stress coping. A theory driven randomized field experiment was used to test a model of stress and coping with job loss. The programatic hypothesis was that vigilant coping and social support are needed for making and implementing difficult decisions such as those following job loss. These elements were combined to generate an intervention protocol. The experimental treatment included training in vigilant coping during job search and provided specific forms of social support hypothesized to promote internalization of vigilant coping patterns. The experimental condition added peer support in the form of pairs of employed participants who telephone each other frequently to discuss their efforts during the intervention program. The experiment included outcome measures such as job seeking and intervening variables including social support and social undermining which are linked by a process model. The longitudinal panel design that surrounds the intervention allowed the testing of a number of hypotheses concerning the effects of the intervention and other variables on mental health, and on active coping with unemployment through job seeking behavior. Although data collection is not complete, it is hoped principles of intervention useful for coping with stress will be derived. (ABL)