The optimal number of plasma exchanges in Guillain-Barré syndrome had not been established. Recently, a clinical trial by the French Cooperative Group showed that 2 exchanges were sufficient for mild case of Guillain-Barré syndrome, but all others 4 exchanges were optimal. The supplementary examination is needed to validate the conclusion, but a certain period should be taken. As the second best way to determine the appropriate number of plasma exchanges, we simply examined the reduction rate of IgG on each session. Wilcoxon signed-ranks test showed that removal rates of IgG differed between the first and second sessions, and between the second and third sessions. At least, 2 exchanges seemed to be needed for Guillain-Barré syndrome. Our results confirmed that IgG is the most relevant factor in the pathogenesis of the syndrome.