In Sibrant et al. (2014), we reconstructed the evolution of Graciosa Island (Azores). We extensively discussed the nature and the meaning of the destruction episodes, either tectonics or gravitational, and concluded that the island has evolved through major landslides. In their comment, Quartau et al. (2015) conclude that (1) “Sibrant et al. (2014) is based almost solely on subaerial observations,” which is false because we used the bathymetric data available to us (Figs. 3 and 4 in Sibrant et al., 2014). (2) “…the published multibeam sonar data around Graciosa reveals that their proposed successive phases of destruction of the volcanic edifices composing the island by massive landslides is incompatible with the high-resolution bathymetry.” First, saying that the data were published is misleading because only two images are now provided in Quartau et al. (2015). Most of the high-resolution data used by Quartau et al. (2015) are not published, and they still do not release the data for us to analyse and use in this reply. Second, the high-resolution bathymetric maps are not incompatible with our model. For instance, mounds on the eastern submarine slope may actually be landslide blocks, and the platform developed to the S may correspond to flank collapses of the successive volcanoes, blanketed more recently by the young basaltic cover. (3) “The interpretation of collapse structures appears to have originated partly from a misreading of the volcano-stratigraphy and tectonic structures”. We certainly did not “misread” the volcanic stratigraphy and tectonic structures in Graciosa; in great contrast to Quartau et al. (2015), we (a) used major unconformities to establish the volcano-stratigraphy, (b) calibrated this stratigraphy with high precision K–Ar dating, (c) made careful measurement of lava flow attitudes to infer the pre-collapse position of the main edifices, and (d) did not use hypothetical tectonic faults, as Quartau et al. (2015) did, to draw an evolutionary cartoon of Graciosa. The lava flows in the southern cliffs of Graciosa dip inland, as recognised by Gaspar (1996, Fig. 3B), in great contrast to the claims of Quartau et al. (2015), and therefore do not support successive volcanoes coinciding with the centre of the island. (4) From unreleased high-resolution bathymetric data, Quartau et al. (2015) consider that no large landslide(s) occurred in Graciosa, and propose a “new model.” First, their geological model is not new, as it reproduces a cartoon in Gaspar (1996, Fig. 3A). Second, Quartau et al. (2015) have not reported any major faults exposed at the surface, as expected for an island-scale graben, but their summary cartoon of Graciosa evolution is based on major faults. Third, the new model of Quartau et al. (their Fig. 8, and our Fig. 4) misrepresents the initial model of Gaspar (1996, Figs. 4.1 and 4.2), as the whole SW flank of the island has been arbitrarily removed on their drawing (see our Fig. 3). In contrast, Sibrant et al. (2014) propose that the SW flank has been effectively removed, but by landslides. Therefore, their new model for the evolution of Graciosa is based on misleading hypotheses rather than convincing arguments. In conclusion, there is no need to change our proposed new stratigraphy and evolution of the island based on Quartau et al.'s claims because these are not supported by actual geological data—well established and calibrated stratigraphy, geometry of lava flows, and geometry and kinematics of faults. Given the absence of sound geological data to support the evolutionary cartoon proposed by Quartau et al. (2015), we conclude that it is mostly flawed and therefore mere speculation.