The moderating influence of perceived social support and sociability on the relationship between hostility and physiological coronary heart disease risk factors was studied in a randomly selected sample of 671 healthy 18 to 30-year-old young adults, It was found that, when hostility, perceived social support, and sociability were treated as separate factors, they operated as suggested by the previous studies, i.e., social support and sociability retained their positive role and hostility its negative one (social support and sociability were related to a low level of physiological risk parameters while hostility predicted a high physiological risk level, correlations being, however, not very high). Their interactions were, instead, slightly unexpected. Social support was likely to suppress the morbid effect of hostility in women, but not in men. This was especially true of anger, and discussed in terms of social desirability and gender-related differences of anger expressions. Sociability was responsible of the most surprising interaction: high sociability increased the harmful effect of hostility, both in men and in women. It was suggested that sociability does not guarantee social skills, but only provides a high frequency of social contacts. If hostility prevents mature social learning, the combination of high hostility and high sociability is likely to lead to an increased CHD risk.