One hundred patients, eighty-nine men and eleven women, with chronic stable angina who were previously selected for aortocoronary bypass grafting gave informed consent for non-invasive and invasive testing of hemodynamic responses to symptom-limited maximal exercise before surgery. Psychosocial coping strategies were evaluated preoperatively by structured interviews and assessment of patients perceptions of symptoms (Cornell Medical Index) and life changes (Holmes and Rahe Schedule of Recent Experiences). Preoperatively forty-one patients were "compartmentalized," forty-two "generalized" and seventeen "vacillated" according to Josten's classification of coping strategies. The Berle Index of social assets was lower and the prevalence of psychiatric symptoms (Cornell categories M to R) was greater in the vacillators preoperatively. Despite less ischemic ST depression in vacillators, no other significant physiological differences were noted between these categories preoperatively. Postoperatively more of the vacillators refused follow-up evaluation, and of vacillators who returned, only one-half were adequately revascularized at operation. Of sixty-five reevaluated after surgery, eight improved, twelve worsened and forty-five did not change classification of coping strategies, yet physiological variables of cardiac function when invasively measured in sixty patients were significantly improved in all three groups. Amounts of improvement, both absolutely and relative to sex- and age-adjusted normal values, were least in vacillators with virtually normal cardiac capacity, and/or inadequate revascularization. Compartmentalized patients were more frequently working, yet only sixty-four in all psychosocial classifications worked before surgery. After this event only forty-five resumed working; none of the non-workers or retired returned to work. Both physiologic improvement and working status were independent of postoperative psychosocial status.