Introduction Although much attention has been paid to a positive association between obstructive sleep apnea (OSA) and impaired glucose metabolism, no information is available on the effect of continuous positive airway pressure (CPAP) on plasma glucose levels in oral 75-g glucose tolerance tests (75-g OGTTs) in patients with OSA. Materials and methods A total of 47 non-diabetic patients who underwent full polysomnography for suspected OSA and who were newly diagnosed as having OSA with an apnea–hypopnea index (AHI) ⩾20 were prospectively enrolled in this study. The patients were divided into the following 2 groups: 29 patients in whom CPAP could be introduced (CPAP group) and 18 patients in whom CPAP could not be introduced (non-CPAP group). All patients underwent 75-g OGTTs at baseline and after 6months. Results There were no significant differences in age, body mass index, prevalence of men, hypertension, dyslipidemia or current smokers, medications and polysomnographic parameters between the 2 groups. In the CPAP group, AHI decreased significantly (51.8±15.7–5.6±5.5, p<0.001). Plasma glucose levels at 2h after glucose load decreased significantly after 6months in the CPAP group (149.0_} 27.8mg/dl to 132.9±27.1mg/dl, p=0.009), whereas the levels did not change significantly in the non-CPAP group (141.7±28.9–151.8±46.9mg/dl, p=0.33). The body mass index, fasting plasma glucose level and homeostasis model assessment for insulin resistance did not change in the both groups. Conclusion This study indicates that CPAP therapy can decrease the plasma glucose level after glucose load in non-diabetic patients with OSA, suggesting that CPAP therapy has a favorable effect on glucose metabolism even in non-diabetic patients with OSA.