The starting point for the presented study was the concept by House who construed social support as buffering the impact of work-related stress on health. Self-esteem was taken under consideration as the other potential stress buffer. It was hypothesized that both social support and self-esteem would have a salutogenic effect, since they attenuate the experience of occupational stress and reduce health problems associated with the experienced job stress. Participants in the study were 361 medical professionals representing various specialties. They were examined using the Subjective Job Evaluation Questionnaire by Dudek et al., the Mood and Health State Questionnaire by Rzasa, the Self-Esteem Scale by M. Rosenberg and Significant Other Scale by Power et al. The higher was the respondents' occupational stress, the poorer was their subjective physical health. Such components of occupational stress as responsibility, psychological strain due to job complexity, lack of rewards at work, and a sense of threat were found to be most important in this respect. These four components of occupational stress were interrelated and constituted a feedback loop. The study confirmed a salutogenic role of self-esteem, contributing to subjective health improvement. Satisfaction with social support had also a positive role, since it reduced the amount of experienced job stress, thus exerting a health-promoting effect. There was a direct negative feedback loop between self-esteem and somatic health problems. Irrespective of that, satisfaction with social support was found to interact with perceived occupational stress in a negative feedback loop. However, neither of these two factors, i.e. self-esteem and social support, had an effect of buffering the impact of occupational stress on health. This suggests that the initial model proposed by House as well as the present author's earlier research findings obtained from a smaller sample should be revised.