This study is a qualitative analysis and a quantitative analysis of all peripheral biofeedback client data files of the University of North Texas Biofeedback Research and Training Laboratory since its establishment in 1991 and through the year of 2002. The purpose of this study is to evaluate the clinical and educational efficacy of the BRTL. Clients' electromyography and temperature measures, self-report of homework relaxation exercises and progress, and the pre- and post-Stress Signal Checklist were reviewed and analyzed. In regard to clinical efficacy, results indicate statistically significant changes in both temperature training and muscle tension training as a whole group. When divided into subtypes based on the clients' primary presenting problem, findings indicate statistical significance in chronic pain, tension headache, and temporomandibular jaw pain on temperature training, and show statistical significance in chronic pain, tension headache, hypertension, migraine headache, stress, and temporomandibular jaw pain on muscle tension training. When analyzing the Stress Signal Checklist, only 25% of clients had complete information on both pre- and post-Stress Signal Checklist. For these 25%, 87.5% reported symptoms decreased. When reviewing the clients' self-reported progress in therapist's session notes, there is no procedure for computing a treatment success to failure ratio due to the inconsistency of therapists in recording clients' statements. This study also identifies three basic biofeedback learning curves that show how people learn self-regulation skills in biofeedback therapy: 1) steady state and trainable (low variability), 2) phasic state and trainable (high variability), and 3) phasic state and low trainable (high variability).