An experiment investigated the mechanisms by which humans estimate Euclidean distances on the basis of kinaesthetic cues. Blindfolded participants followed straight and curvilinear paths with a hand-held stylus (encoding phase). Then, with a straight movement, they estimated the Euclidean distance between the start-and end-points of the path (response phase). The experiment contrasted an On-axis condition, in which encoding and response movements were spatially aligned, and an Offaxis condition, in which they were displaced laterally. Performances were slightly more accurate in the On-axis condition than in the Off-axis condition. In both conditions, however, errors were consistently smaller when the path covered a larger surface. The results showed that small paths yielded an overestimation of the Euclidean distance, the relative errors increasing with the length of curvilinear paths. The findings are compared with results of other studies in which distances were estimated on the basis of haptic cues.