Affiliation and employment histories obtained from a sample of skid-row men and a control sample of residents in a low-income metropolitan census tract are used in a test of the hypotheses that skid-row men are less affiliated than lower-class men in settled neighborhoods, and that downward occupational mobility is associated with loss of affiliations. Compared with the control sample, skid-row men have long histories of low affiliation, both before and after their arrival on skid row. The two samples did not differ much in occupational mobility, but their affiliative patterns have been quite different. Apparently the disaffiliation of skid-row men cannot be attributed to their downward mobility. Whether downward mobility is accompanied by disaffiliation seems to depend on the context in which it occurs.