This report focuses on the question 'Why does stimulation of a few hairs of a cat's unsupported paw lead to biceps activation and contact placing (CP) while stimulation of the supported paw does not?' The rubral system is used as the model; individual rubral neurons (RN) are recorded in the awake cat during CP and rubrospinal tract neurons (RTN) are identified by demonstrating antidromic invasion. Classes of RN were identified corresponding to each of the main stages of CP. In particular, a contingent, contact-locked category was identified, responding only if placing followed the contact. Another category, the pre-delayed movement-locked, responded 40-160 ms prior to delayed placing. Both of these categories imply that the rubral system dynamically participates (with other higher motor neurons) in CP, a conclusion strengthened by the positive relationship between RN discharge and movement parameters under varying load conditions. Physiological stimulation of the supported paw may fail to excite RN, which respond strongly to gentle contact of the unsupported paw (which leads to CP). Cooling or making permanent lesions in nucleus interpositus and dentatus drastically reduces resting and evoked RN and especially RTN activities in the awake cat. Some evidence of recovery was present by 4-6 weeks after the lesions. The effect of behavioral context was also studied on the biceps responses in the awake cat to a train of 3-4 pulses at 500 Hz delivered either to the red nucleus or to the rubrospinal tract (RT). The biceps responses were enhanced when the forepaw was removed from a support (the pre-condition for CP).(ABSTRACT TRUNCATED AT 250 WORDS)