Fifty working subjects in the ages 25 to 60 were asked to rate the amount of job strain (demands divided by decision latitude) that they experienced on four occasions at intervals of three to four months. Blood samples were taken for the analysis of immunoglobulin G before work started on these work days. After the end of the studied year each subject was asked to rate the total amount of social support that he/she had experienced. Both availability and adequacy of support were rated. The reported degree of job strain was unrelated to season and number of order of observation. Immunoglobulin G was observed to rise progressively with rising job strain. The reported adequacy of social support was negatively associated with immunoglobulin G--the more adequate the support was reported to be, the lower the immunoglobulin G level. This association was stronger at peak levels of job strain and became weak and nonsignificant at low levels of job strain. Thus, when job strain increased immunoglobulin G increased mainly among those with poor adequacy of social support.