Because of their unstable pathophysiology, it was hypothesized that patients with non-Q-wave acute myocardial infarctions (AMI) would be more vulnerable to the negative effects of psychological stress than patients with Q-wave AMI, and thus would be more likely to benefit from programs aimed at relieving stress. This hypothesis was tested through secondary analysis of data from a 1-year randomized clinical trial of psychological stress monitoring and intervention after AMI. After discharge, treatment group patients were telephoned each month and asked to respond to an index of psychological stress symptoms (General Health Questionnaire GHQ-20). Those with high stress symptoms (GHQ greater than or equal to 5) received home nursing visits. Control group patients received usual care. The sample consisted of 461 men, aged 31 to 86 years, who responded to the GHQ-20 before hospital discharge. Patients were followed for 5 years using record data. There were 321 Q-wave AMIs, 112 non-Q-wave AMIs and 28 indeterminate electrocardiograms. Life-table analyses showed that among patients with non-Q-wave AMIs receiving usual care, high stress in the hospital (GHQ greater than or equal to 5) was associated with a 1-year relative risk (RR) of cardiac mortality of 5.49 +/- 1.39 (p = 0.01). In comparison, control patients with Q-wave MIs had no stress-related increase in risk (RR = 0.41 +/- 2.08, p = 0.40). In the treatment group, the patients with non-Q-wave AMIs did not experience an increase in risk associated with high stress (RR = 1.80 +/- 1.79, p = 0.52).(ABSTRACT TRUNCATED AT 250 WORDS)