Hand muscle reflexes following muscle stretch and electrical nerve stimulation show a typical pattern consisting of short- and long-latency reflexes. The present investigation was designed to test reflexes following pure cutaneous stimulation. Air puffs were delivered to the palmar tip and the nail bed of the first, second and fifth fingers during isotonic contraction of hand muscles. The EMGs from the thenar muscles, the first dorsal interosseous muscle and the hypothenar muscles were recorded. Reflexes were obtained in all muscles, with a typical configuration consisting of a short-latency excitatory component (cutaneous long-latency reflex I, cLLR I) and a second excitatory component (cutaneous long-latency reflex II, cLLR II), with an inhibitory component between them. The size of cLLR II differed depending on the area stimulated and the muscle recorded. We found the largest responses always in the muscle acting on the stimulated finger. The reflex size depended on the strength of air puff stimulation. Allowing small displacements of the fingers led to an additional increase in the size of the reflex. The pattern of reflexes was identical independent of whether the finger tip or the nail bed was stimulated, but the size of the reflexes was smaller following nail bed stimulation. Following blockade of the cutaneous nerve branches of the thumb with local anaesthetics, air puff stimulation of the thumb no longer elicited this reflex pattern. Hence, under our experimental conditions, cutaneous receptors were the only source of afferent input for these reflexes.(ABSTRACT TRUNCATED AT 250 WORDS)