Twenty-five in-depth interviews were made with middleaged persons of both sexes, who had suffered a first myocardial infarction (MI) two months previously. The purpose was to assess the layman's understanding of concepts such as social network, social support and control which are exposures commonly used or suggested for use in quantitative research in the area of social epidemiology. The validity of the instruments and the underlying concepts in assessing such exposures has important implications for the interpretation of the association between psychosocial factors and health or health behavior. The layman's understanding of social network and social support concepts seemed unproblematic, but the concept of control was understood in rather disparate ways among those interviewed. The experience of control/decision latitude in the work environment, seemed to form a model for the overall understanding of control among several of those interviewed and there also seemed to be important differences based on gender. Another impression, was that the individual's level of aspiration might be an important confounder in the assessment of control in population studies. It is therefore suggested that epidemiologcal data should be stratifyed by basic demographic variables like age, sex and social class in analyses including the control variable, to improve the valid use of this type of measure.