Gender differences in causal attributions and emotions for imagined success and failure on examinations were investigated. Males made stronger ability attributions for success than females, whereas females emphasized the importance of studying and paying attention. Males more than females attributed failure to a lack of studying and low interest, but females were more likely than males to blame an F on a lack of ability. Females experienced stronger emotions than did males; they felt happier than males did after success but felt more like a failure than did males after imagining receiving an F on an examination. Some of the gender differences in causal attributions, especially for ability attributions, depended on the gender-type of the subject matter of the examinations. The implications of these findings are discussed.Causal attributions for achievement have been studied extensively over the past twenty-five years. Weiner (1974) originally categorized causal attributions along two dimensions: stability and locus of control. Although Weiner (1985) added a third dimension, controllability, most research on gender differences in causal attributions has focused on his two original dimensions. The attributions for achievement outcomes that have received the most empirical attention include ability (internal, stable), effort (internal, unstable), task difficulty (external, stable), and luck (external, unstable). Research has found that, in general, people show a self-enhancing bias in attributions; participants attribute greater responsibility to ability and effort in explaining their successes than failures, whereas failure participants attribute greater responsibility to external factors such as task difficulty than do success participants (e.g., Arkin & Maruyama, 1979;Elig & Frieze, 1979;Gilmor & Reid, 1979).In 1979, Falbo and Beck discovered a methodological problem with most of the research on causal attributions of performance: Only 23 percent of the causal attributions participants made spontaneously "could be classified in terms of the Weiner et al. (1971) model. Effort constituted 13% of the total; Ability, 8%; Task Difficulty, 1%; and Luck, less than 1%" (p. 188). Which causal attributions constitute the other 77 percent of responses is an intriguing question. To address this issue, the present research allowed pilot test participants to list the causes of receiving an A or an F in one of three subject matters. These free responses were then used in a closed-ended format on a separate set of participants. It was hoped that this approach would yield information on the kinds of causal attributions people make spontaneously.