Multiple sclerosis (MS) is primarily a disease of the central nervous system. Although the involvement of the peripheral nervous system in MS was suggested over 100 years ago, the issue is still controversial, and it is generally accepted that except for the optic nerve the peripheral nerves are left unaffected by the disease. We hypothesize, that an electroneurographical study if thorough enough, may reveal differences in some nerve conduction parameters between MS patients and healthy subjects. Second, we assume that the sensitivity of nerve conduction measurements might be increased if performed at a range of temperatures, reflecting a differential effect of cooling and warming on the peripheral nerve conduction parameters in MS patients and controls. Finally, we expect that the differences in these parameters between controls and MS patients will increase with the progression of the disease. To test these hypotheses in a pilot study, we performed a detailed analysis of the motor and sensory nerve conduction features of the right median nerve in 13 MS patients and 13 controls at 5 degrees C increments between 20 and 40 degrees C, and repeated these measurements after 3 years. The motor latencies were 0.3-0.6 ms longer in MS patients compared to the controls both initially and 3 years later (0.058<p<0.09). The durations and areas of the compound motor action potential (CMAP) appeared more sensitive to changes in temperature in the MS group (0.057<p<0.1). The change in both distal motor latency and sensory latency per unit change in temperature decreased significantly in 3 years within the MS but not in the control group. These results suggest a mild and progressive involvement of the PNS in MS. Most differences in this pilot study were on the border of statistical significance therefore our hypotheses should be confirmed in studies with larger sample size.