Muscle strength in neuropathic patients is usually evaluated clinically using manual muscle testing (MMT). Detection and grading of mild symmetrical muscle weakness using MMT is difficult partly because the examiner must take into consideration the normal variation in strength in relation to age, weight, height, and gender. In the present study assessment of the strength of ankle dorsal and plantar flexors and knee flexors and extensors with MMT and isokinetic dynamometry were compared in 108 patients, of whom 86 bad diabetes mellitus and 22 had alcoholic liver cirrhosis. The isokinetic muscle strength of the patients was compared with the strength of 90 healthy control subjects, adjusted for the influence of age, weight, and height for both genders. MMT resulted in a significant underestimation of the frequency and severity of muscle weakness in both the ankle and the knee. In 28-41% of the comparisons, MMT misclassified the strength performance with one category or more (>25%). Misclassifications were most frequent for the ankle plantar flexors.