This report examines the accuracy of subjective appraisals of several aspects of task performance, including proficiency, difficulty, frequency, and criticality. An introduction discusses current Army use of subjective appraisal, feedback methods, and problems with subjective appraisal. Data pertaining to the accuracy of various types of appraisal are summarized in the next section. The types of appraisal included are proficiency appraisals, task criticality appraisals, task difficulty appraisals, task frequency estimates, and appraisal of training materials. At the end of the section, research from the cognitive psychology literature relating to human ability to make accurate subjective appraisals is discussed. The third section summarizes data relating to the relative accuracy of appraisals obtained from alternative sources, namely supervisor, self-, and peer appraisals. In the fourth section, discussion of the issue of how subjective appraisals are formulated centers around survey and interview techniques. Methods for increasing the accuracy of subjective appraisals are made, focusing on phrasing of questions, raters' experience, and other characteristics of raters. The paper concludes with suggestions for optimizing combined use of the survey and interview approaches. (YLB)