1. Simultaneous recordings were made from postganglionic sympathetic fibres supplying hindlimb skin and skeletal muscle in chloralose-anaesthetized, artificially ventilated cats. Single-fibre activity was either isolated by dissection or discriminated from few-fibre preparations of fascicles in the left superficial peroneal or sural nerve (innervating hairy skin) and common peroneal nerve (innervating muscle). Vasoconstrictor fibres were identified by their spontaneous activity as well as their responses to stimulation of the lumbar sympathetic chain and to changes in baroreceptor activity. The baroreceptors were then denervated by bilateral section of the vagi, carotid sinus and aortic nerves. 2. In five cats, neurones in the region of the subretrofacial nucleus were activated chemically by microinjections of 2-10 nl 0.5 M-sodium glutamate from a micropipette inserted into the ventral surface of the medulla. Both skin and muscle vasoconstrictor fibres were activated by glutamate injections into this region on either side of the medulla. Arterial pressure also rose. 3. Glutamate injections at forty-two sites evoked a positive response, defined as an increase in cutaneous and/or muscle vasoconstrictor fibre activity of at least 25%. This response was evoked only in the cutaneous fibre at sixteen of these sites ('skin points'), only in the muscle fibre at seven sites ('muscle points'), and in both fibres in the remainder ('mixed points'). The largest percentage increases in activity of either type of fibre were obtained from mixed points. 4. The blood pressure rises following glutamate stimulation of muscle points were significantly greater than those produced by stimulation of skin points. Analysis of all positive responses showed that the evoked rise in blood pressure was significantly correlated with muscle sympathetic activity but not with cutaneous sympathetic activity. 5. Glutamate stimulation at different sites could evoke differential responses in skin and muscle vasoconstrictor fibres without any detectable change in the pattern of phrenic nerve discharge. 6. Skin points were grouped in the medial part of the subretrofacial region, and muscle points in the lateral part. In addition, for all positive responses there was a highly significant correlation between the ratio of muscle to cutaneous sympathetic activity evoked, and the distance from the mid-line of the corresponding injection site. 7. These results demonstrate a functional differentiation among subretrofacial neurones in their relative control of the sympathetic vasoconstrictor supply to skin and skeletal muscle.(ABSTRACT TRUNCATED AT 400 WORDS)