1. Reciprocal inhibition between antagonist muscle groups at the ankle has been investigated in sixty healthy subjects. Hoffmann reflexes (H reflexes) in the soleus and tibialis anterior muscles were used to assess changes in reciprocal inhibition evoked by electrical stimulation of antagonist muscle nerves. 2. Inhibition of the soleus H reflex was evoked by a single conditioning stimulus to the common peroneal nerve, and inhibition of the tibialis anterior H reflex was elicited by one conditioning stimulus to the posterior tibial nerve. Symmetrical central connections between the antagonist flexors and extensors were assumed and under this assumption the central delay for the inhibition, in addition to the delay for monosynaptic Ia excitation, was calculated to be about 1 ms. The inhibition was evoked by weak stimuli to the nerves from antagonist muscle groups; the threshold for the inhibition was around 0.6 X threshold for a direct motor response (M-threshold). Furthermore, tendon taps to the Achilles tendon facilitated the soleus H reflex and inhibited the tibialis anterior reflex at short latencies. The short central delay, the low electrical threshold and the.actions of Achilles tendon taps strongly suggest that the early reciprocal inhibition is homologous to the disynaptic Ia inhibition previously studied in animal experiments. 3. With the test soleus H reflex kept at 15-25% of the maximum directly evoked motor response (M-response) and the strength of the conditioning peroneal nerve stimulation kept at 1.0 X M-threshold, the inhibition from the peroneal nerve ranged between 0 and 40% (mean, 14.9%) at rest. 4. Changes in the amount of reciprocal inhibition from the peroneal nerve were studied both during tonic and dynamic dorsi- and plantarflexion. During tonic dorsiflexion there was no significant change of inhibition as compared to rest, while inhibition decreased during tonic plantarflexion. However, during ramp-and-hold dorsiflexion there was a transient increase in reciprocal inhibition of the soleus H reflex. This increase in inhibition from the peroneal nerve could be seen 50 ms prior to the onset of contraction. The increase in inhibition before and at the very beginning of the contraction cannot be due to sensory feed-back during contraction, but must depend on a supraspinal control of the spinal cord. 5. At conditioning-test intervals of 4-6 ms, the inhibition of the soleus H reflex from the peroneal nerve was considerably larger during tonic dorsiflexion than at rest. Thus, tonic dorsiflexion revealed an inhibition with long latency from the peroneal nerve, which was not seen at rest.(ABSTRACT TRUNCATED AT 400 WORDS)