Non-depressed individuals exhibit a self-serving attributional bias, taking more credit for success than for failure. Clinically and subclinically depressed people are less self-serving, often to the point of making similar attributions (explanations) for successes and failures. The present studies evaluated a schematic account of these distinct attributional biases. Subclinically depressed and non-depressed participants completed measures of attributional bias (the relative strength of ability attributions for success versus failure), schemabased optimism (the relative expectedness of success versus failure) and self-schemas of competence. Two studies evaluated a hypothesis derived from the schematic account: the greater the perceived competence and optimism, the more self-serving the attributional bias. As predicted, (a) attributional bias scores covaried with optimism and competence scores in both magnitude and valence (or direction), (b) depressed-non-depressed differences in attributional biases paralleled differences in competence and optimism and (c) when attributional bias scores were adjusted for the effects of optimism or competence, depressed-non-depressed differences in attributional biases were eliminated. The schematic account raises questions about the common assumptions that attributional patterns are traits, and that they play a central role in the aetiology and treatment of depression.