To examine a self-efficacy explanation of the finding that depression is related to hemoglobin A1c (A1c) level in people with type 1 but not type 2 diabetes.Cross-sectional design involving 124 participants with type 1 (n=32) and type 2 (n=92) diabetes. Participants completed measures of depression and diabetes-related self-efficacy. A1c was obtained from medical records.Replicating prior findings, A1c was significantly correlated with depression in type 1 participants (r=.51, p<.01), but not in type 2 participants (r=.11, ns). As hypothesized, A1c was significantly correlated with self-efficacy among type 1 participants (r=−.42, p<.05) but not among type 2 participants (r=−.01, ns). Self-efficacy also mediated the effect of A1c on depression among the type 1 participants (Z=2.21, p<.05).In people with type 1, but not type 2 diabetes, A1c levels are related to diabetes adherence mastery (self-efficacy), which mediates the link between A1c and depression. Results are discussed with regard to the proposal that perceptions of ineffective control over one's health play a role in the development of depression (a consequence model of depression in diabetes).