Study objective To evaluate the tolerability and to find predictors of treatment success for an individually adjusted, one-piece mandibular advancement device in patients with snoring and obstructive sleep apnea. Design Prospective study. Setting Departments of Respiratory Medicine and Orthodontics, Umeå University. Patients Six hundred nineteen of 630 patients (98%), who consecutively received treatment for sleep apnea and snoring from February 1989 to August 2000, were followed up. They had a mean apnea-hypopnea index of 16 (range, 0.0 to 76) and a mean body mass index of 28 (range, 19 to 42). Measurements Interviews, questionnaires, and overnight sleep apnea recordings. Patients with an apnea-hypopnea index of ≥ 10 in the supine and/or lateral position were considered to have obstructive sleep apnea. A lateral apnea-hypopnea index of < 10, together with a supine apnea-hypopnea index of ≥ 10, defined supine-dependent sleep apneas. Results One hundred forty-eight of the 619 patients (24%) discontinued treatment. Female gender predicted treatment success, defined as an apnea-hypopnea index of < 10 in both the supine and lateral positions, with an odds ratio of 2.4 (p = 0.01). In the women, the odds ratios for treatment success were 12 for mild sleep apnea (p = 0.04), and 0.1 for complaints of nasal obstruction (p = 0.03). In the men, the odds ratios for treatment success were 6.0 for supine-dependent sleep apneas (p < 0.001), 2.5 for mild sleep apnea (p = 0.04), 1.3 for each millimeter of mandibular advancement (p = 0.03), and 0.8 for each kilogram of weight increase (p = 0.001). Conclusions The mandibular advancement device is recommended for women with sleep apnea, for men with supine-dependent sleep apneas defined by a lateral apnea-hypopnea index of < 10, and for snorers without sleep apnea. Men who increase in weight during treatment reduce their chance of treatment success and are advised to be followed up with a new sleep apnea recording with the device. To evaluate the tolerability and to find predictors of treatment success for an individually adjusted, one-piece mandibular advancement device in patients with snoring and obstructive sleep apnea. Prospective study. Departments of Respiratory Medicine and Orthodontics, Umeå University. Six hundred nineteen of 630 patients (98%), who consecutively received treatment for sleep apnea and snoring from February 1989 to August 2000, were followed up. They had a mean apnea-hypopnea index of 16 (range, 0.0 to 76) and a mean body mass index of 28 (range, 19 to 42). Interviews, questionnaires, and overnight sleep apnea recordings. Patients with an apnea-hypopnea index of ≥ 10 in the supine and/or lateral position were considered to have obstructive sleep apnea. A lateral apnea-hypopnea index of < 10, together with a supine apnea-hypopnea index of ≥ 10, defined supine-dependent sleep apneas. One hundred forty-eight of the 619 patients (24%) discontinued treatment. Female gender predicted treatment success, defined as an apnea-hypopnea index of < 10 in both the supine and lateral positions, with an odds ratio of 2.4 (p = 0.01). In the women, the odds ratios for treatment success were 12 for mild sleep apnea (p = 0.04), and 0.1 for complaints of nasal obstruction (p = 0.03). In the men, the odds ratios for treatment success were 6.0 for supine-dependent sleep apneas (p < 0.001), 2.5 for mild sleep apnea (p = 0.04), 1.3 for each millimeter of mandibular advancement (p = 0.03), and 0.8 for each kilogram of weight increase (p = 0.001). The mandibular advancement device is recommended for women with sleep apnea, for men with supine-dependent sleep apneas defined by a lateral apnea-hypopnea index of < 10, and for snorers without sleep apnea. Men who increase in weight during treatment reduce their chance of treatment success and are advised to be followed up with a new sleep apnea recording with the device.