1. 1. Experimental evidence indicates that the onset of CNS pathologic changes occurs in the preparalytic period and is closely associated with the earliest evidence of virus activity in any particular region involved. 2. 2. The earliest cytopathologic changes are diffuse chromatolysis of Nissl substance in the cytoplasm of nerve cells and mild cellular exudate consisting of polymorphonuclear and mononuclear leukocytes. 3. 3. Nerve cell changes may be present in the earliest stages without inflammatory reaction in the vicinity and therefore are not necessarily the result of the latter, but rather the result of direct virus action. 4. 4. Nerve cell changes either lead to rapid destruction of the cell or to arrest in the stage of cytoplasmic chromatolysis, following which complete morphologic recovery of the cell generally occurs over a period of about a month or less, depending upon the severity of injury. 5. 5. Virus activity, nerve cell changes and inflammatory reaction are localized only in certain susceptible regions of the CNS, largely due to specific differences of susceptibility of nerve cells. The intensity of the inflammatory reaction, however, may be quite variable in different susceptible centers and in different individuals. Severe inflammatory reaction is usually but not always associated with extensive nerve cell destruction. Severe nerve cell damage may occur without extensive cellular infiltration in the cord. 6. 6. Lesions in the cerebral cortex are usually confined to the motor area of the precentral gyrus and even here the lesions are rarely severe enough to suggest that they may produce clinical symptoms. 7. 7. “Encephalitic” symptoms such as restlessness, stupor, disorientation and coma are associated with severe inflammatory reaction in the brainstem and often with small softenings in this region. They are not associated with unusual involvement of the cerebral cortex. 8. 8. Brainstem centers principally involved in most instances are the reticular formation of the hind-brain, the vestibular nuclei and the roof nuclei of the cerebellum. Resulting functional disturbances are discussed. 9. 9. Widespread dissemination of virus among most motor nerve cells in spinal cord enlargements occurs in experimental poliomyelitis as early as the first day of paralysis. Motor nerve cells which are affected either are destroyed very quickly during the first few days of the disease or undergo slower recovery changes leading to complete morphologic recovery within about a month. After this time it can be shown that the degree of paralysis and atrophy are closely correlated with the number of motor nerve cells destroyed. In the acute stage, however, this correlation is not as high and other factors must also play a role in producing paralysis. An important factor is the reversible injury of motor nerve cells. Less complete evidence from human material suggests that a similar situation obtains in human poliomyelitis. 10. 10. Experimental work suggests three possible factors which may determine the variation in severity of infection. These are, first, variations due to difference in strains of the virus, second, reduction of severity due to previous paralytic or non-paralytic infection, and third, host variation unrelated to previous immunizing experience with the virus.