Three experiments involving different angular orientations of tactual shapes were performed. In experiment 1 subjects were timed as they made 'same-different' judgments about two successive rotated shapes. Results showed that no rotation effect is obtained, i.e., reaction times and error percentage do not increase linearly with rotation angle. The same negative results were found in experiment 2, in which subjects were similarly timed while they made mirror-image discriminations. In experiment 3 a single-stimulus paradigm was used and subjects were asked to decide if a rotated stimulus was a 'normal' or 'reversed' version. Reaction times increased linearly with angular departure from the vertical. Therefore, for tactual stimuli too, this study confirms previous results, which suggest that a mental rotation strategy only occurs if it is facilitated by both type of task and type of stimulus. Results also show a significant difference between hands, and between hands and type of response. Implied hemispheric differences are discussed.