1. The possibility that Ia afferent fibres from the gastrocnemius medialis (GM) and from the tibialis anterior (TA) muscle could converge on to a single interneuronal pool inhibitory to the soleus motoneurones was investigated. 2. The soleus H reflex, evoked by tibial nerve stimulation in the popliteal fossa, was conditioned by separate or combined stimulation of the nerves to the GM or TA muscles. Stimulus intensity was below the motor threshold (MTh), and the conditioning-test intervals were such as to evoke short-latency inhibition of the soleus H reflex. Care was taken to avoid current spread and artifacts connected with the closeness in time and space of the conditioning and test stimuli. 3. Separate stimulation of both GM and TA nerves was able to induce significant inhibitory effects on the H reflex amplitude at stimulus strengths larger than 0.75 x MTh, on the average. Combined stimulation of the two nerves was able to reduce the H reflex at lower stimulus strengths, at which either nerve was ineffective alone. 4. Conditioning stimulus strengths close to the MTh reduced the H reflex to approximately 80% of the control value, both on single and combined stimulation, i.e. saturation of the inhibitory effect was found. 5. By extrapolating the regression line through the normalized data from all subjects, it was assumed that the smallest stimulus strength necessary to drive the inhibitory interneurones to threshold was, on the average, 0.5 and 0.6 x MTh, on combined and separate nerve stimulation, respectively. 6. Tonic voluntary activation of the soleus abolished the inhibitory effects of both separate and combined stimulations. This was tested on the H reflex, on the rectified and averaged EMG, and on the peristimulus histogram of single motor unit discharge. 7. The findings strongly suggest the existence of spatial summation of the effects from GM and TA muscle at the level of a single interneuronal pool. Most probably, the responsible afferent fibres are group I spindle afferents, and the interneurones those mediating the reciprocal inhibition. The data do not support the notion of parallel pathways, exclusive to each nerve.