Background and objective: Around 25% of patients with neuro-muscular diseases (NMD) are treated by home noninvasive ventilation (NIV) through an oronasal mask. However, there is growing evidence that nasal masks require lower NIV pressures and result in fewer residual obstructive events. We hypothesized that nasal masks would improve efficacy and reduce side effects compared to oronasal masks in this population. Methods: open label, cross-over, randomized, study in 2 tertiary care hospitals. Patients with NMD treated by home NIV were randomized for one-week periods to nasal and oronasal interfaces respectively (cross-over). At the end of each period, nocturnal polygraphy (monitoring mouth opening) under NIV, synchronized with transcutaneous partial pressure in CO2 (tcCO2) was performed. Data were collected from the NIV built-in software and NIV side-effects were collected. Intention-to-treat and per protocol analyses were performed. The primary outcome was mean nocturnal SpO2. The secondary outcomes were: percentage of sleep with SpO2 < 90%, oxygen desaturation index (ODI), mean tcCO2, mean duration of mouth opening during sleep, level of non-intentional leaks and side-effects. Results: Thirty patients with NMD were included. There were no between-group differences for either the primary or secondary outcomes. Post hoc comparisons showed that changing between interfaces reduced NIV efficacy: mean nocturnal SpO2 (p = 0.04), ODI (p = 0.01), mean tcCO2 (p = 0.048), side-effects (p = 0.008). Conclusion: Nasal masks did not improve NIV efficacy or reduce side effects compared to oronasal masks in patients with NMD treated by home NIV. The efficacy of NIV is reduced during the transition to another interface, requiring close monitoring. Registration number: NCT03458507.