Client distortion of test scores was investigated under conditions of immediate and delayed recall. The subjects were 59 college students. Procedures included test administrations, individual test‐interpretation interviews, recall of scores, and several methods of profile analysis. Results indicated a high degree of accuracy on all measures of immediate recall; marked variability and significant distortions (p < 0.001) appeared on delayed recall. The chief characteristics of distortion for delayed recall were overestimation (p < 0.03) and leveling (p < 0.01). Owing to the nature of distortion on delayed recall, the use of several methods of profile analysis was supported in individual cases.