Mouthpiece ventilation (MPV) allows patients with neuromuscular disease to receive daytime support from a portable ventilator, which they can disconnect at will, for example, for speaking, eating, swallowing, and coughing. However, MPV carries a risk of underventilation. Our purpose here was to evaluate the effectiveness of daytime MPV under real-life conditions. Eight wheelchair-bound patients who used MPVunderwent daytime polygraphy at home with recordings of airflow, mouthpiece pressure, thoracic and abdominal movements, peripheral capillary oxygen saturation (SpO2), and transcutaneous partial pressure of carbon dioxide (PtcCO2). Times and durations of tasks and activities were recorded. The Apnea-Hypopnea Index (AHI) was computed. Patient-ventilator disconnections ≥3 minutes and episodes of hypoventilation defined as PtcCO2>45 mmHg were counted. Patient-ventilator asynchrony events were analyzed. The AHI was >5 hour(-1) in two patients. Another patient experienced unexplained 3% drops in arterial oxygen saturations at a frequency of 70 hour(-1). Patient-ventilator disconnections ≥3 minutes occurred in seven of eight patients and were consistently associated with decreases in SpO2 and ≥5-mmHg increases in PtcCO2; PtcCO2 rose above 45 mmHg in two patients during these disconnections. The most common type of patient-ventilator asynchrony was ineffective effort. This study confirms that MPV can be effective as long as the patient remains connected to the mouthpiece. However, transient arterial oxygen desaturation and hypercapnia due to disconnection from the ventilator may occur, without inducing unpleasant sensations in the patients. Therefore, an external warning system based on a minimal acceptable value of minute ventilation would probably be useful.