This study investigated the effects of enhanced perception of control on adherence to a worksite relaxation program. The design was a 2 x 2 between subjects manipulation in which desire for control was an individual difference measure and condition was an experimenter assigned treatment. Subjects were randomly assigned to either Perceived Control (PC) condition or a Standard Program (SP) format. The perception of control manipulation was operationalized as the opportunity for important choices in the program. It was hypothesized that subjects in the perceived control condition would adhere significantly better than those in the standard program condition. An interaction between motivation for control and condition was predicted such that subjects high in desire for control would adhere best in the PC condition, and the converse would hold for those low in motivation for control. Sixty university staff and faculty who volunteered for a relaxation training program were offered an orientation and three training sessions, and were given the opportunity of attending four practice sessions. Dependent measures of adherence included attendance and self-monitoring of home relaxation practice. Dependent variables of secondary interest included pre-post measures of anxiety, physical symptoms and perception of control over health. Analysis of the adherence measures demonstrated a significant effect of condition on attendance at training sessions only, p < .0045, accounted for by the great number of drop-outs in the SP condition. There were no significant differences between conditions on other adherence measures. There was no statistically significant effect of condition on dropping out, per se. A ceiling effect on attendance by those subjects who continued in the program obscured any other differences. There was no evidence of an interaction between condition and desire for control on adherence. / Master of Science