This study describes the prevalence of alcohol-related problems and de
velops predictive models to explain racial differences in subsamples o
f 494 black and 568 white men from a national probability survey of dr
inking patterns and problems. The results showed that although black m
en exhibited higher mean scores on many types of alcohol-related probl
ems, they did not report significantly higher rates of heavier drinkin
g and drunkenness. nor did they score higher on a scale of permissiven
ess of drinking norms. A hierarchical regression analysis indicated th
at race independently predicts problem scores even when controlling fo
r other social and demographic factors. Moreover. an interactive model
showed that race interacts significantly with the frequency of heavie
r drinking and some sociodemographic characteristics. As the frequency
of heavier drinking increases, rates of drinking problems rise faster
among black men than white men. Religion and unemployment also had di
fferent effects on rates of alcohol-related problems in each group of
men. These findings suggest that racial differences in the prevalence
of drinking problems might be related to differences in the sociocultu
ral context of drinking and in the material conditions under which bla
ck and white men live.