In this study, experimenters (pseudo-coaches) provided feedback that varied in valence, sequence, and amount to 50 male college students. A laboratory analogue paradigm was used that included a basketball-like underhand free throw task in which subjects first were instructed on proper technique and then took 10 baseline shots (trials) followed by 2 blocks of 20 trials each. Subjects were randomly assigned. Some interacted with a pseudo-coach who made no comments during the two experimental trial blocks (control), while others received feedback (6-8 comments per trial block) that was response-specific, emotionally oriented, and provided in one of four sequences: positive-positive, negative-negative, positive-negative, or negative-positive. Based on prior research on coach behavior and social psychological studies of interpersonal behavior, we hypothesized that both of the continuous feedback groups would show performance decrements and associated reactions to the coach and the task. These predictions were supported regarding performance and, to some extent, regarding a measure of sustained self-observation. Discussion includes interpretation of the nominally superior performance of the control group, the nonsignificant results on the subjective evaluation measures, and implications of these findings in view of external validity criteria and prior analyses in the emerging behavioral technology of coaching.