The prevalence of Charcot‐Marie‐Tooth's disease (CMT) was studied in Western Norway, an area with several isolated districts with a population of 725,000 (1968). Three hereditary types were distinguished in the area: autosomal dominant CMT with an estimated prevalence of 36/100,000; X‐linked recessive CMT with a prevalence of 3.6/100,000; and autosomal recessive CMT with a prevalence of 1.4/100,00. Gene frequencies were 3 · 7. 10^‐4, 1 · 9. 10^‐4, and 4 · 8. 10^‐4 in autosomal dominant, X‐linked, and autosomal recessive CMT, respectively, while the corresponding mutation rates were 13 · 0, 5 · 5, and 3 · 5 per million gametes per generation. The penetrance was almost complete for all three variants of CMT.

Strict diagnostic criteria were followed in the selection of the 37 index cases. A family investigation was carried out with 238 subjects, during which 69 secondary cases were detected. Another 57 subjects had unspecific neuropathy (Un), which did not fit a diagnosis of CMT or other neurological disease. In the diagnosis of Un, a score system was used, with age and sex corrections based on findings in a normal population.

Generally, the most severe disease course was found in the recessive CMT types, but there was also more clinical variation, suggesting CNS involvement in some cases (upper motor neuron affection, cerebellar signs). Scoliosis and spinal ataxia were not infrequent, even in cases with autosomal dominant CMT.

The prevalence cf Un was highest in the relatives of recessive CMT cases, with a ratio of affected to normal in sibs compatible with a hypothesis of several cases of heterozygous manifestation. In the relatives of autosomal dominant CMT cases, Un prevalence was also higher than in the population, but lower in 2nd degree relatives than in 1st degree; the ratios fitted a hypothesis of polygenic Un inheritance. Significant differences were found in the score patterns of Un in the recessive CMT families and in the autosomal dominant families, suggesting their difference of origin. The reason for clustering of Un cases in autosomal dominant CMT families is obscure, since it can be only partly attributed to early manifestation of CMT. It is suggested that Un and CMT, mainly in autosomal dominant CMT, interact to form a spectrum of differing phenotypes, so explaining the problem of “transitional forms” between CMT and other hereditary nervous disorders. Recessive CMT, being a more generalized nervous disease, attains, through differing expressivity, phenotypes which vary between individual cases.