Background. Most studies of social support appear to assume that it is something that the external environment provides to an individual. However, longitudinal and genetic studies are beginning to question this assumption. The purpose of the present study is to differentiate quantitatively the intra-individual determinants of social support from its extra-individual components. Methods. The subjects were 103 high school exchange students who were enrolled in a 1-year placement with a host family in various countries of the world. The People in Your Life scale, a reliable and validated self-report measure of perceived social support, was administered before their departure from home, after 6 months of stay in a completely unfamiliar foreign community and 6 months after their return home. Structural equation models were examined that partitioned the intra-individual and extra-individual components of social support, and the best fitting models were selected. Results. Between 24% and 69%, or up to 86% depending on the situation, of the variances of social support measures were stable across situations and considered intra-individual; 31% to 76% of what is measured as social support was of extra-individual origin. Moderately strong correlations were noted between the intra-individual components and the trait extraversion. Conclusions. Measures of social support, which is usually tacitly understood as something extrinsic to an individual, in fact may not be measures of the external environment only.