A single study investigated the effects of discrimination on performance. One hundred and three female undergraduates completed an initial anagram test which they believed to be a scholastic aptitude test. They were then given either an overt, ambiuous, or no cue regarding the likelihood that the evaluator discriminated against women. They were later told that a second session of a similar task would be evaluated by the same or a different evaluator. All participants were informed that they had performed poorly on the first task. Results indicated that although the women tended to attribute the failure to external factors (discrimination) rather than internal ones (ability and effort) as discrimination cues increased in strength, only those who received ambiguous discrimination cues did not improve performance from the first to the second task. Implications for the effects of ambiguity on performance are discussed.