The dynamic process of how non-invasive positive pressure ventilation (NPPV) improves spontaneous ventilation is unclear. Therefore, daytime trends of blood gases and breathing pattern were assessed by measurements 0, 0.5, 1, 3, 7, 11 and 15 h after cessation of nocturnal controlled NPPV in patients with chronic hypercapnic respiratory failure. Twelve patients (six COPD/six restrictive) who were established on NPPV and 12 controls (six COPD/six restrictive) completed. PaCO2 decreased during controlled NPPV (P < 0.02). PaCO2 additionally decreased step by step during the first 3 h of spontaneous breathing after switching from NPPV to spontaneous breathing (P < 0.05), but remained unchanged in controls. The PaCO2 decrease was due to a stepwise increase in tidal volume (P < 0.05) at an unchanged breathing frequency. Accordingly, minute ventilation also stepwise increased (P < 0.03). There were no significant changes in controls. Improvements of PaCO2 and tidal volume occurred even after 5.7 +/- 3.1 days following first NPPV trials, but became more evident after 2 months. Maximal inspiratory mouth pressures increased in chronic obstructive pulmonary disease (COPD) patients (P < 0.05), and respiratory drive increased in restrictive patients (P < 0.05) following 2 months of NPPV. Lung function parameters and inspiratory impedance remained unchanged. Improvements in health-related quality of life were evident and were correlated to the decline of elevated bicarbonate levels (r = 0.72, P < 0.01). In conclusion, there is a stepwise adaptation process lasting 3h when switching from nocturnal controlled NPPV to daytime spontaneous breathing in which tidal volume increases and PaCO2 drops after an initial PaCO2 decrease while on NPPV.