fifty community-dwelling adults aged 51-83 years of age (m=70.1, sd=7.61) reported their medication compliance and their use of memory strategies. the first goal of the present study was to replicate gould mcdonald-miszczak, and king’s (1997) research on older adults’ use of memory strategies to aid medication compliance. as in the gould et al. (1997) study, older adults reported using internal strategies more often than external strategies for medication compliance, and the use of strategies was predicted by metamemorial variables (p<.05) rather than by objective health related factors. the second goal of this study was to extend the research by gould et al. (1997), who found that metamemorial variables, rather than objective medical factors, were significant predictors of self-reported medication compliance. the present study used a new selfreport measure of compliance, and examined cognitive performance in addition to health related and metamemorial assessments. the self-report measure of compliance was tied more to objective health factors than to subjective beliefs, as medical factors contributed significantly to the prediction of self-reported compliance (p=.05). finally, a metacognitive path model specific to self-reported compliance was tested, and the importance of compliance, a belief laden variable, predicted self-reported compliance through prospective strategy use. the results of this study suggest that medication compliance is a complex mixture of both objective health related factors and subjective metacognitive assessments.