A new roughness perception test was designed to evaluate cutaneous sensibility. Blindfolded subjects explored a sample of sandpaper using one of two methods: stroking the sample (dynamic method), or applying pressure (static method). A range of samples of various grades were presented in a random order, and subjects scored each one in terms of perceived roughness. Each subject performed the test in three conditions -wearing latex and nitrile examination gloves and bare-handed. Mean normalised scores for each combination of sandpaper grade and hand condition were calculated. The COF between each sample and glove (or bare finger) was measured, and the topography of each sample was analysed using a profilometer. It was found that the COF did not vary significantly across the samples, and so could not be related to perceived roughness. However, there were strong correlations between perceived roughness and surface topography (roughness average, particle diameter, particle spacing), particularly in the dynamic method. In the static method, most subjects did not perceive roughness differences below 800 µm particle spacing. In both methods, there was a significant reduction in perceived roughness when gloves were donned, but no significant difference between the two glove types could be found. It was concluded that the dynamic method was a useful test for evaluating cutaneous sensibility. Further investigation of the relationship between friction and perceived roughness was recommended, using a wider range of sample materials.