The major purposes of this cross sectional study were to investigate the mediating and interactive mediatingeffects of received social support and various dimensions of health locus of control on the relationship between perceived stress and anxiety. The sample consisted of 67 subjects who had experienced a first-timemyocardial infarction within the past three to six weeks. The data were collected using Cohen's Perceived Stress Scale, Spielberger's State-Trait Anxiety Inventory Form Y - I, Barrera's Inventory of Socially Supportive Behavior and Wallston's Multidimensional Health Locus of Control Scale.A significant positive relationship between received social support and internal health locus of control was found when the sample was analyzed as a whole. When subgroup analyses were performed, a mediating effectof received social support was found for the subjects with higher socioeconomic status and higher educational background but not for those with lower socioeconomic or educational status. This indicates that received social support was beneficial for those under high stress, but did not seem to make difference for those under low stress. Further analyses showed that high internal health locus of control (IHLC) was favorable for males with low stress, but not for those with high stress. Powerful others health locus of control(PHLC) has an interactive mediating effect with received social support for the older subjects, but not the younger ones.When the types of received social support were tested for mediating and interactive mediating effects, it was found that social interaction and chance health locus of control (CHLC) and tangible support and PHLC each had an interactive mediating effect on the relationship between perceived stress and anxiety.These findings suggest that there is a relationship between received social support and internal health locus ofcontrol (IHLC), but the role of IHLC may be different for males than for females. Furthermore, older persons with low PHLC may be better able to use received social support. Those with low PHLC may benefit most from tangiblesupport while those with high CHLC may benefit most from the social interaction type of support in high stress situations.Finally, the data from this study suggest that an interactive model might be better to understand the relationships between stressor, stress, and anxiety than a simple model.