Social capital is often claimed to be promoted by stable social structures such as low migration rates between neighbourhoods and social networks that remain stable over time. However, stable social structures may also inhibit the formation of social capital in the form of social networks and social participation. One example is psychosocial conditions at work, which may be determined by characteristics such as demand and control in the work situation. The study examines the active workforce subpopulation within the Swedish Malmö Shoulder Neck Study. A total of 7836 individuals aged 45-69 years, were interviewed at baseline between 1992 and 1994, and at a 1-year follow-up. Four groups of baseline psychosocial work conditions categories defined by the Karasek-Theorell model (jobstrain, passive, active, relaxed) were analysed according to 13 different social participation items during the past year reported at the 1-year follow-up. Odds ratios and 95% confidence intervals with the jobstrain group as a reference were estimated. A multivariate logistic regression model was used to assess differences in different aspects of social participation between the four psychosocial work conditions groups. The results show that the respondents within the active category in particular but also the relaxed category, have significantly higher participation in many of the 13 social participation items, even after multivariate adjustments. The results strongly suggest that psychosocial work conditions may be an important determinant of social capital measured as social participation, a finding of immediate public health relevance because of the well known positive association between social participation and health-related behaviours.