If the groups of the Eoe classification system were unrelated to real categories in the occupational structure, the groups would be random subsamples of the total sample. Measures for three kinds of stability were denned. The frequencies of job changes within the group and to other groups were compared with each other and with the total sample by computing chi-squares. The groups were found to differ significantly from each other with respect to the measures of stability. When people are grouped according to this classification, their occupational behavior is not random and is associated in some way, direct or indirect, with the group to which they belong.