Treatment options for obstructive sleep apnea (OSA) may involve potential side effects or discomfort; nasal continuous positive airway pressure (CPAP) may not be tolerated by 25% of patients. We therefore sought to determine the efficacy of mandibular advancement as a treatment for OSA, and to investigate whether clinical and radiographic parameters can predict the response to this treatment. Sixteen male and 3 female subjects with documented OSA who had failed or been unable to tolerate nasal CPAP underwent baseline polysomnography and cephalometry, and were then fitted with a removable Herbst appliance to achieve forward mandibular advancement during sleep. All subjects then underwent a second cephalometric evaluation and polysomnography with the appliance in place. Fourteen of 15 subjects demonstrated significant improvement in the degree of OSA, based on the apnea-hypopnia index (AHI) (34.7 +/- 5.3 to 12.9 +/- 2.4 events/h, p < 0.002). Comparison of pre- and posttreatment cephalometric values revealed no significant change in the posterior airway space (PAS) despite a reduction in mean AHI. There was a significant decrease in the mandible-hyoid distance (MP-H) with treatment for the group as a whole. When the study population was evaluated on the basis of a successful response to mandibular advancement (posttreatment AHI < 10), the baseline MP-H was found to be significantly shorter in the responders than in nonresponders. MP-H after mandibular advancement was likewise shorter in responders than in nonresponders. In addition, the soft palate length (PNS-P) showed a significantly greater shortening in responders after treatment.(ABSTRACT TRUNCATED AT 250 WORDS)