H-reflex recovery by twin pulses was recorded serially in 10 paraplegics for 5 months and in 23 hemiplegics for 2 months after the lesion. Fifty-one normal subjects acted as controls. The effect of cutaneous tactile stimulation was also studied simultaneously by applying electrical stimuli synchronized with twin pulses to the skin over the lateral border of small toe. In paraplegics, the H-reflex recovery curves recorded serially showed a highly depressed pattern during the first two weeks, an almost normal pattern during the second and third months and a significantly elevated pattern during the fourth and fifth months. Whereas cutaneous stimulation in control subjects produced a highly significant late inhibition of H-reflex recovery between 600 ms and 600 ms, in paraplegics it failed to produce any significant effect, except in two, who besides having a normal H-reflex recovery curve even during the first week, showed a substantial amount of cutaneous inhibition of H-reflex recovery, 4 months after the lesion. A highly depressed pattern of H-reflex recovery was observed on the affected side of the majority of hemiplegics during the first week after the lesion, many of them showing similar pattern on the "unaffected side" also. The serial study showed very good improvement in all hemiplegics both in terms of H-reflex recovery pattern and the amount of cutaneous inhibition. The observations in present study suggest preservation and/or restoration of supraspinal influences in many hemiplegics and in at least two paraplegics. The study also shows that a serial recording of H-reflex recovery curve and the amount of cutaneous reflex effect on it, is a very sensitive method of assessing the supraspinal influences on the spinal motoneurones and so can be of immense help in the diagnosis and prognosis in hemiplegics and paraplegics.