Using a novel trial design, we prospectively examined the effect of intravenous immunoglobulin in seven patients with chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) in a double-blind, placebo-controlled cross-over study. We suggest that the commonly used manual muscle testing and Rankin scale are not sufficiently sensitive to measure changes in CIDP and should not be used as isolated outcome measures. We propose a timed 10-m walk, the Nine-Hole Peg Test, the Hammersmith Motor Ability Score, and myometry as alternative measures which are valid, reliable and sensitive. Our trial design permitted the measurement of a treatment response in three responders despite different patterns of disability typical of the broad clinical picture seen in CIDP.