Purpose

To study the internal consistency, factorial structure, and convergent validity of the Swedish translation of the General Self-Efficacy scale (S-GSE). Methods

The S-GSE and two items on mental and physical work capacity were completed by a randomized population cohort (n = 4,027) and two cohorts (n = 3,310 and n = 498) of incident cases of sick-leave (>14 days). Results

S-GSE means were higher among men than women in two of the cohorts and higher in the randomized population cohort than in the two sick-leave cohorts. Internal consistency was high with α = .90. Unidimensionality was indicated and factor loadings ranged between .64 and .80. Moderate correlations (.35–.38) between the S-GSE and mental work capacity were found in all cohorts. Yet, the correlation between S-GSE and physical work capacity was weaker in the sick-leave cohorts. The psychometric properties showed similar patterns across gender. Conclusions

Across three cohorts, additionally stratified by gender, the S-GSE comprised one single latent factor and showed high internal consistency. However, since S-GSE was more strongly related to self-assessments of mental work capacity than physical work capacity regardless of sick-leave status, the S-GSE may not be a strong predictor of beliefs about physical work capacity across all populations.