This cross-sectional, correlational study examined the relationships of diabetes-specific treatment barriers and self-efficacy with self-care behaviors. A total of 309 people with type 2 diabetes participated in this study. All of the factors were assessed by self-report questionnaires. Self-care behaviors included exercise, diet, skipping medication, testing blood for glucose, adjusting insulin to avoid or correct hyperglycemia, and adjusting diet to avoid or correct hypoglycemia. Perceived barriers to carrying out self-care behaviors were associated with worse diet and exercise behavior. Greater self-efficacy predicted more frequent blood glucose testing, less frequent skipping of medication and binge eating, and closer adherence to an ideal diet. Nontraditional dimensions of self-efficacy were associated with worse self-care. Self-efficacy explained 4% to 10% of the variance in diabetes self-care behaviors beyond that accounted for by patient characteristics and health beliefs about barriers. The findings of this study provided support for Rosenstock's proposal that a person's self-perceived capability to carry out a behavior should be incorporated into an expanded health belief model.