1. (1) The sympathetic skin response (SSR) is a slow wave, generated in deep layers of the skin, resulting from reflex activation of the sudomotor sympathetic efferent fibres. The aim of this study was to define experimental conditions, best stimulation and recording procedures, and the criteria for validation of the responses. 2. (2) Thirty normal subjects (aged 25-56) were tested. The stimulation was an electrical pulse train applied to the median nerve at the wrist, a binaural tone burst, or both simultaneously. Records were made with surface electrodes on hand and foot contralateral to the stimulated median nerve. 3. (3) Response shape was most often biphasic in feet, biphasic or triphasic in hands. SSR amplitude was 3.1 ± 1.8 mV in hands, 1.4 ± 0.8 mV in feet. Normal mean onset latency was 1.5 ± 0.08 sec for hand response, 2.05 ± 0.10 sec for foot response. The mean conduction velovity along peripheral sympathetic nerve fibres was 1.40 ± 0.14 m/sec in lower limbs. 4. (4) Bimodal stimulation (burst + median) provided responses of larger amplitude. The influence of stimulation intensity was also investigated. A decrease in amplitude and lengthening of latencies were observed after 15-20 min of testing. 5. (5) The criteria for validation of responses are discussed. The importance of central processing time in the response delay is pointed out. In good methodological conditions, SSR would appear to be a simple, effective means of assessing sympathetic sudomotor outflow in central and peripheral nervous system disorders. © 1990.