Understanding the challenges faced by organisms moving within their environment is essential to comprehending the evolution of locomotor morphology and habitat use. Geckos have developed adhesive toe pads that enable exploitation of a wide range of microhabitats. These toe pads, and their adhesive mechanisms, have typically been studied using a range of artificial substrates, usually significantly smoother than those available in nature. Although these studies have been fundamental in understanding the mechanisms of attachment in geckos, it is unclear whether gecko attachment simply gradually declines with increased roughness as some researchers have suggested, or whether the interaction between the gekkotan adhesive system and surface roughness produces nonlinear relationships. To understand ecological challenges faced in their natural habitats, it is essential to use test surfaces that are more like surfaces used by geckos in nature. We tested gecko shear force (i.e., frictional force) generation as a measure of clinging performance on three artificial substrates. We selected substrates that exhibit microtopographies with peak‐to‐valley heights similar to those of substrates used in nature, to investigate performance on a range of smooth surfaces (glass), and fine‐grained (fine sandpaper) to rough (coarse sandpaper). We found that shear force did not decline monotonically with roughness, but varied nonlinearly among substrates. Clinging performance was greater on glass and coarse sandpaper than on fine sandpaper, and clinging performance was not significantly different between glass and coarse sandpaper. Our results demonstrate that performance on different substrates varies, probably depending on the underlying mechanisms of the adhesive apparatus in geckos.