The abilities of humans to discriminate the softness of rubber objects of differing compliance with a hand-held tool (a stylus) was measured under experimental conditions that differed as to how the tool was used and the kind of sensory information available. When the subject actively tapped or pressed the compliant objects, they discriminated softness as well by means of a stylus as they did by contacting the objects directly with the fingerpad. Discrimination with the stylus was unaffected by whether the stylus was controlled by one or two fingers. While tapping or pressing a stylus held in a precision grip, the grip force increased before, reached a maximum at the same time as, and decreased in parallel with the compressional force. This relationship was suggestive of anticipatory motor control based on an internal model of the motor system and the physical properties of the object. Discrimination was significantly better when tapping as opposed to pressing the objects with the stylus. This was hypothesized as due to the presence of tactile cues generated by the rapid increase in force rate as the stylus struck and indented the object during tapping. During tapping, the magnitude and rate of compressional force produced by the stylus against the object were greater, the harder the object. An additional cue, possibly kinesthetic, during pressing and tapping was the magnitude of indentation of the specimen by the stylus that was greater, the softer the object. Subjects could discriminate differences on softness by tactile cues alone in the absence of kinesthetic when compliant objects were tapped at approximately the same velocity by the experimenter against a stylus in contact with the subject's passive fingerpad. Discrimination deteriorated if the softer specimen of a pair was tapped with a slightly greater velocity than the harder and not possible if the specimens were pressed against the stylus without generating tactile cues of mechanical contact. In contrast, discrimination was possible during active pressing and unaffected by variations in velocity during active tapping. It is concluded that during active movements, kinesthetic information and knowledge of central efferent commands provide useful cues that are not present during passive touch. These cues allow the observer to discriminate differences in object compliance not confounded by differences in applied velocity.