The purpose of the present study was to evaluate the long-term efficacy of oral appliances (OAs) in early treated patients with obstructive sleep apnea (OSA). Polysomnographic sleep recordings without and with an OA were performed at treatment start and in patients who had been continuously treated with OAs for at least 15 years. Nine patients (eight men) with a median age of 68.1 years (interquartile range (IQR) 60.0 to 76.3 years) and a median treatment time of 16.5 years (IQR 16.3 to 18.0 years) were included. The apnea-hypopnea index decreased from a median of 17.3 (IQR 9.7 to 26.5) to 7.2 (IQR 4.0 to 9.6; p = 0.03) at the short-term follow-up. After long-term use, the apnea-hypopnea index was 32.4 (IQR 22.2 to 58.8) without the device and 35.1 (IQR 13.6 to 46.2) with it (p = 0.08). There were increases in the apnea-hypopnea index, both without the device (p = 0.02) and with it (p = 0.008). The degree of mandibular advancement did not differ between the two study occasions (p = 1.0). Patients treated with oral appliances may experience deteriorations in disease severity and treatment efficacy during continuous long-term OA treatment. Regular follow-up schedules with renewed sleep apnea recordings should be considered for these patients in order to avoid suboptimal or a total loss of effects on sleep apneas.