Considering many psychosocial health risk factors are interrelated, determining psychosocial health risk might benefit from a more person-centered perspective. This paper explores to what extent a psychosocial profile that combines potentially synergistic effects of different psychosocial characteristics, including psychological attributes and functioning, coping styles and social support, predicts self-rated health, morbidity and mortality. Prospective, longitudinal data from 1,912 Dutch participants aged 55-91 years were used to determine distinct psychosocial profiles by means of two-step cluster analysis. The predictive power of these profiles over a 5-year follow-up was calculated with Cox regression models for all-cause mortality and general practitioner-diagnosed somatic morbidity, and logistic regression models for self-rated health. Three distinct psychosocial risk profiles emerged: an adverse, an average and a beneficial profile. These profiles strongly predicted self-rated health but not morbidity or mortality. The health effects of the cluster (profile) model suggest synergism between the psychosocial characteristics. Future research should replicate our findings to further validate the approach.