The experiment tested the effect of verbal praise and reproof on performance on a balance board similar in principle to that described by Reynolds (1952). Eight female college students tested 48 college males and 48 college females and administered standard verbal comments of praise to one-half the Ss (randomly designated) and reproof to the remaining Ss. Each examiner tested 12 Ss on three separate occasions, and Ss performed six 30-sec. trials on each occasion. Four examiners tested male Ss, and four tested female Ss. The estimate of internal consistency for the mean of six trials for males was .91 and .94 for females. Ss were initially told that the testing was being done to establish norms, but the true purpose was made known on completion of the experiment. Analysis of covariance, computed by analysis of variance of residual gain scores, yielded no significant reinforcement (praise and reproof) effect. The mean of two nonreinforced trials was used in computing residual gain scores. There was a significant (# < .01) Reinforcement-by-Examiner interaction. Reproof administered by one personable and physically attractive examiner greatly improved performance, but reproof administered by another, who was not personable and was plain in appearance, decreased performance. Examiners had practiced a standard testing procedure prior to the experiment and minimized extraneous conversation with Ss during the experiment. There were no other observable differences of consequence among these two examiners and the six remaining examiners. The confounding of reinforcement with examiner appearance was unplanned. It is possible that the effect of praise and reproof is dependent on examiner characteristics and sociocultural gender factors. The significant Reinforcement-by-Examiner interaction emphasized the need for including an examiner factor in reinforcement studies. Hakstian and Brickner (1968) have reported an Investigator-by-Reinforcement interaction in a verbal-conditioning experiment. The absence of an examiner factor in other studies may, in part, explain apparently conflicting reports on effects of praise and reproof.