Objective: The present study examined self-efficacy and self-esteem as basic aspects of the self that influence self-care and physiological outcomes among young adults with Type I diabetes. The two aims of this study were 1) to examine the cross-sectional and longitudinal role of the self-variables as they predict self-care and HbA1c and 2) to test whether self-care mediates the association between the self variables and HbA1c using cross-sectional and longitudinal data. Methods: One hundred ten participants were recruited from a regional diabetes outpatient clinic. Inclusion criteria were age (18-35 years) and duration of diabetes (>1 year before recruitment). Participants were 61% female and 88% white. In addition, the sample had an average annual income between $24,999 and $34,999, and 85% had completed some or all of college. The average duration of diabetes was 15 years. Results: Using multiple regression analyses we found that, compared with self-esteem, self-efficacy was a better predictor of all aspects of self-care and HbA1c in cross-sectional analyses, in addition to diet and exercise self-care, and a better predictor of HbA1c in longitudinal analyses. The data also supported the cross-sectional and longitudinal mediational model in which better self-care helped account for the association between greater self-efficacy and better HbA1c. Conclusions: Self-efficacy is an important factor for management of self-care practices and physiological outcomes among young adults with Type I diabetes, and self-care may be an important mechanism by which self-efficacy influences HbA1c levels.