We would like to thank Drs. Esquinas, Ucar, and Kirakli for their interest in our paper, and for highlighting the importance of positive airway pressure (PAP) mask design. Before discussing the interesting points raised by Esquinas et al., there are two important issues that we wish to clarify. Firstly, the authors raised their confusion that changing from a nasal to an oronasal mask resulted in increased leak and residual apnea-hypopnea index (AHI) without a concurrent change in therapeutic pressure requirement. We reported a statistically significant—although not clinically relevant—difference in residual AHI across the three mask types when the PAP device was set to continuous mode (i.e., fixed pressure); when the device was set to automatically-adjusting PAP (APAP), there was no difference in residual AHI between the masks. In the absence of a difference in the frequency of respiratory events, one would not expect a difference in delivered pressure. Secondly, Esquinas et al. suggested that we had not considered the fact that differences in the exhalation port of each mask will lead to differences in leak and pressure requirement. As discussed in the methods section of our paper, however, we strongly agree with the authors that the exhalation port specific to each mask design is relevant both in terms of differentiating unintentional versus intentional leak, and in terms of the ability of the PAP device to apply an appropriate therapeutic pressure. As Esquinas et al. have highlighted in their letter, several aspects of mask design probably play a role in allowing the APAP device to deliver an appropriate level of pressure on a breath-by-breath basis. These include the size and design of the exhalation port, the dead space, the surface area of the seal, and the effect that the mask has on the position of the mandible. In our clinical study, we chose to focus on the latter by including measurements of incisal relationship and mandibular protrusion. Although the other design features listed above are no less important, they are best addressed via bench studies of the nature that Esquinas et al. have listed [1–3] rather than a clinical trial. We chose to conduct the majority of data collection in patients’ homes and were therefore reliant on the ResMed S9 device to detect leaks as well as any ongoing respiratory events (technology that we are confident is adequate for this purpose [4]). We agree with Esquinas et al. that a limitation of this study design is that we were not able to address potential differences in sleep efficiency/arousals or measurements of functional outcomes or sleepiness; however, we suggest that a broader clinical trial could easily address this limitation. Finally, we understand Esquinas et al.’s concern that by recruiting patients who were established nasal continuous positive airway pressure (CPAP) mask users, we were not able to address nasal pathology. We would emphasize, however, that this issue is unavoidable if the study aim is to compare nasal and oronasal CPAP masks in a clinical population. If investigators were to recruit all comers without excluding those unable to tolerate the use of a nasal mask, this arm of the study would yield little meaningful data. We thank Esquinas et al. for their letter, and we look forward to reading the results of further bench studies and clinical trials in this area. J. P. Bakker (*) Division of Sleep Medicine, Brigham and Women’s Hospital, Harvard Medical School, 221 Longwood Avenue Suit 036BLI, Boston, MA 02113, USA e-mail: jpbakker@partners.org