Being held accountable for one's views in an ambiguous situation has been found to increase cognitive processing and reduce possible biases often pervasive in performance appraisals (Tetlock, 1983a). The present study examined the effects of making raters more accountable for their performance ratings through expectations of a supervisory review. Two types of accountability were manipulated. Interpersonal accountability induced subjects to believe that they would receive feedback based on their justifications of their ratings. Reward accountability subjects expected that they would receive a reward based on the quality of their ratings compared to true performance scores. Half of the subjects were also told that the supervisor gave the ratee a negative performance rating. 191 female subjects rated an interviewer's performance after receiving accountability and view manipulations. Results indicated that subjects who received accountability manipulations in conjunction with the view gave more unfavorable ratings and were less accurate on Cronbach's (1955) elevation component of accuracy than no accountability and/or no view groups. Accountable subjects who did not receive the view were expected to increase cognitive processing. These hypotheses pertaining to increases in cognitive complexity and dimensional accuracy were not supported. Although unexpected, interpersonal accountability in conjunction with reward accountability prevented a decrement in processing when given the view. A biased versus unbiased vigilant processing hypothesis was proposed to explain these results. Possible limitations are also proposed as well as implications for future research.