1. Adult owl monkeys were trained to detect differences in the frequency of a tactile flutter-vibration stimulus above a 20-Hz standard. All stimuli were delivered to a constant skin site restricted to a small part of a segment of one finger. The frequency-difference discrimination performance of all but one of these monkeys improved progressively with training. 2. The distributed responses of cortical neurons ("maps") of the hand surfaces were defined in detail in somatosensory cortical area 3b. Representations of trained hands were compared with those of the opposite, untrained hand, and to the area 3b representations of hands in a second set of monkeys that were stimulated tactually in the same manner while these monkeys were attending to auditory stimuli (passive stimulation controls). 3. The cortical representations of the trained hands were substantially more complex in topographic detail than the representations of unstimulated hands or of passively stimulated control hands. 4. In all well-trained monkeys the representations of the restricted skin location trained in the behavioral task were significantly (1.5 to greater than 3 times) greater in area than were the representations of equivalent skin locations on control digits. However, the overall extents of the representations of behaviorally stimulated fingers were not larger than those of control fingers in the same hemisphere, or in opposite hemisphere controls. 5. The receptive fields representing the trained skin were significantly larger than receptive fields representing control digits in all but one trained monkey. The largest receptive fields were centered in the zone of representation of the behaviorally engaged skin, but they were not limited to it. Large receptive fields were recorded in a 1- to 2-mm-wide zone in the area 3b maps of trained hands. 6. Receptive-field sizes were also statistically significantly larger on at least one adjacent, untrained digit when compared with the receptive fields recorded on the homologous digit of the opposite hand. 7. There was an increase in the percent overlaps of receptive fields in the cortical zone of representation of the trained skin. A significant number of receptive fields were centered on the behaviorally trained skin site. 8. The effects of increased topographic complexity, increased representation of the trained skin location, increased receptive-field size, and increased receptive-field overlap were not observed in the representations of the untrained hands in these same monkeys. Only modest increases in topographic complexity were recorded in the representations of passively stimulated hands, and no effects on receptive-field size or overlap were noted.(ABSTRACT TRUNCATED AT 400 WORDS)