Summary In rabbits with experimental isoallergic encephalomyelitis, intracutaneous tests with homologous spinal cord suspended in saline gave reactions having the gross and histologic characteristics of the tuberculin response. Skin reactivity could not be passively transferred with large amounts of serum. The development of skin reactivity was correlated with the development of experimental disease, both in degree and in time. Positive corneal reactivity could be demonstrated in approximately one half to one third of the animals developing encephalomyelitis, especially in those surviving longer than 2 weeks. Titration of inoculation dose and skin test dose was carried out. Skin reactivity remained low and encephalomyelitis failed to appear in animals inoculated with low doses of antigen or tubercle bacilli, in animals inoculated subcutaneously or intraperitoneally, in immature animals, and in animals with cachexia. By using the skin testing method as a means of comparing nervous tissue fractions, antigenic effectiveness was found in a residue insoluble in water and ether and not affected by freezing and thawing in the presence of ether. Proteolipid A of beef brain also gave positive reactions. Encephalomyelitis could not be produced by the water soluble, dialyzable fraction described by Hottle et al. This substance was likewise not effective in skin tests. Dicumarol dosage, sufficient to keep the prothrombin level below 50% of normal, had no effect on incidence of experimental disease.