The aim of this study was to scrutinize the relationship between personality type as described by Grossarth-Maticek and Eysenck, health-related behaviours, and indicators of transitory ill health in a community sample. The sample consisted of all per sons aged 40 years (45 men and 35 women) in a Swedish municipality who agreed to take part in a health examination at the primary health care centre of the municipality. The Short Interpersonal Reactions Inventory (Grossarth-Maticek and Eysenck, 1990) was used to measure personality type. Self-report data were obtained regarding seven health-related behaviours. Health was assessed in three ways; self-report (paper and pencil), self-report (interview response to physician), and measures of 36 biological variables including immune system indicators. The allocation of persons to the different personality types proved problematic using the established methods. By combining the types, according to Eysenck's personality model, and performing a cluster analysis on this combination, a ‘healthy’ and a ‘stressed’ profile were identified within both the male and the female group of subjects. Persons in the healthy personality cluster showed more favourable scores on the health-related behaviour indices and on the self-report health scales than the persons in the stressed cluster. They also tended to score lower than those in the stressed cluster on most of the biological markers known to increase during acute stress. The possibility that this implies a higher level of strain for the persons in the stressed cluster on various bodily systems is discussed.