One hundred one normal subjects and 46 patients were investigated. Various objective and subjective tests for polyneuropathies were compared. Motor and sensory nerve conduction velocities are the most sensitive tests. In normals, age hardly influenced nerve conduction velocities. This is believed to be a result of the strict exclusion criteria. Diagnostic sensitivity is also high with the vibration fork test and with vibratometry at the big toe. Results with the method of limits are as reliable and sensitive as more cumbersome techniques, such as the titration method and the forced choice method. Thermal thresholds and cardiovascular tests are less sensitive. The most correct overall classification is attained with a combination of tests reflecting the function of different nerve fiber classes in the peripheral and autonomic nervous systems.