Self-ratings of work performance have been investigated by researchers for a number of years. Previous research has shown that self-ratings are often lenient, inaccurate, and lack convergence with other performance measures. However, self-ratings are less likely to produce a halo effect when compared to other performance appraisal measures. Further, it has been suggested that the inclusion of self-ratings may decrease an employee's defensiveness in the appraisal system. The current study investigated boundary conditions (Type A / B personality type, social comparison information, and public / private rating settings) that might facilitate more accurate, and less lenient self-ratings. Limited support was found for the research hypotheses. A number of theoretical and empirical explanations can be offered to interpret the findings. / Master of Science