The development of a theory of quality of life following serious illness is currently an important field of endeavour in rehabilitation research. The interpersonal relationships of people and elements of the self-concept appear to be salient factors in causal models of subjective well-being. One such model developed to explain characteristic levels of well-being was used in a longitudinal study of adaptation to a first myocardial infarction. Data were collected from a national sample of male cardiac patients and their spouses on the illness and marital situation as predictors of long-term well-being or ill-being. Analyses of data from the first three waves of the study, which is to extend over 5 years after the onset of illness, are to be discussed in the paper. Marital status, the emotional quality of the spouse relationship and long-standing marital stressors were found to have direct and indirect effects on the two dimensions of the Bradburn Affect Balance Scale. The same is true of continuing problems associated with the heart attack relative to perceptions of having coped effectively with the after effects of illness. Differences in self-esteem and personal competence were suggested as mediators of socio-environmental and illness-related influences. The two-factor conception of well-being developed was found to be a useful framework for investigating positive and negative aspects of psychosocial rehabilitation. The same factors that explain differences between happy and unhappy people in social indicators research also appeared to be determinants of different trajectories of adaptation in the wake of a life-threatening illness. Previous research using theoretical models from stress research has overemphasized psychosocial morbidity and stress management and neglected positive processes of adaptation. The identification of love resources related to positive feeling states and life satisfaction has, therefore, not received the attention it deserves. This is especially the case regarding an overemphasis of the concept of social support as a buffer of stress. Instruments were developed to measure high and low marital intimacy, as well as chronic marital role strains, and these measures appeared to explain different trajectories of adjustment to cardiac disease. The two-factor model appears to be useful for future work on quality of life with chronic illness. The same is true of four marital contexts of rehabilitation that were identified in the study.