The aim of this prospective, randomized study was to analyze dental and skeletal side effects after 4 years of treating obstructive sleep apnea (OSA) patients with a mandibular advancement device (MAD) compared with uvulopalatopharyngoplasty (UPPP). With the appliance in position, the mandible was advanced 50% of maximum protrusion capacity (ie, 4-6 mm); the vertical opening between the incisal edges was, on average, 3 mm. Thirty patients in the MAD group and 37 in the UPPP group completed the 4-year follow-up. There were no differences between the MAD and the UPPP groups in any of the dental or skeletal variables measured after the 4-year treatment period. In the MAD group, small but statistically significant changes were found: there was a posterior rotation of the mandible (mandibular line [ML]/nasion-sella line [NSL]) (mean 0.5 degrees [95% confidence interval (CI) 0.1-0.8 degrees ]). Correlated to the posterior rotation of the mandible, the distances incision superius ML, incision superius-NSL, and incision inferius-NSL increased by means (95% CI) of 0.7 (0.5-1.2), 0.8 (0.4-1.1), and 1.3 (0.8-1.8) mm, respectively. Overjet and overbite did not change significantly, nor was there a significant change in the mandibular length. The observed changes were considered clinically insignificant because overbite and overjet stayed within normal limits. Only the vertical position of the maxillary incisors in relation to ML changed to the extent that the 95% CI of the mean for the change was outside that of the mean of the change in the UPPP group and measurement error. Treatment of OSA with a dental appliance is probably a lifelong process, and long-term follow-up studies should therefore be undertaken to control both the treatment effect on OSA and the side effects on the masticatory system.