Hypothesis: Wettability of solid surfaces is mostly probed with sessile drops rather than bubbles because this method is readily followed out. This recurrent use may lead to a misleading connection of certain phenomena to the hydrophobicity/hydrophilicity of materials. For instance, the Cassie-Baxter regime and the wicking effect are generally associated only to hydrophobic and hydrophilic surfaces, respectively. However, the same phenomenology should be observed when air bubbles (underwater conditions) in contact with solid surfaces are used instead. In particular, one might expect that rough-hydrophilic surfaces become superaerophobic due to the appearance of a hybrid dewetting regime, like the Cassie-Baxter regime described for rough-hydrophobic surfaces. Otherwise, rough-hydrophobic surfaces might become superaerophilic due to air-wicking. Experiments: To elucidate this issue, in this work, we analyzed the wettability of surfaces with very different intrinsic contact angle and roughness degree. The analysis was performed with both Sessile Drop and Captive Bubble methods. Findings: Our results with captive bubbles for rough-hydrophilic surfaces revealed phenomena only explained by the occurrence of a transition from the Wenzel regime to an "inverse" Cassie-Baxter regime. In addition, our results with captive bubbles for rough-hydrophobic surfaces showed evidences of air percolation through the interconnected asperities. This effect reminds the wicking effect reproduced on rough-hydrophilic surfaces, responsible for superhydrophilicity. (C) 2018 Elsevier Inc. All rights reserved.