We tested the hypothesis that reducing core body temperature in subjects with multiple sclerosis (MS) improves the cortical somatosensory evoked potential (SEP) response. Twenty subjects with definite MS were compared with 20 subjects without neurologic symptoms or disease. SEPs were recorded with stimulation of the tibial and median nerves unilaterally at 3.1 and 6.1 Hz. The procedure was repeated after a cooking vest and hat reduced core body temperature by an average of 0.46 +/- 0.28 degrees C. No appreciable change in latency or amplitude of the SEP responses occurred in either the control or MS group with cooling. Although the amplitude of the cortical SEP response was less at the 6.1 Hz rate than at 3.1 Hz, there were no statistically significant differences between the MS and control groups or between stimulation rates with cooling. We conclude that, although some reports suggest symptomatic improvements during cooling in subjects with MS, this improvement may not be associated with changes in the SEP.