We have now demonstrated that JC virus interacts uniquely with the human brain as a cause of Progressive multifocal-leukoencephalopathy (PML). The spectrum of viruses causing PML has been uncertain ever since two PML cases were reported to be caused by SV40 virus one year after the isolation of the new polyomavirus JC from PML in 1971. Although the presence of SV40 in PML brain was never confirmed in this country, apparent confirmation came from cases reported from (then) East Germany in 1981 and from Japan in 1985. We have now been able to reexamine those two cases with newer methods not dependent on polyclonal antibodies which may be cross-reactive or low-titered. We have used both in situ hybridization procedures which can distinguish SV40 and JCV DNA and a new monoclonal antibody specific for SV40 capsid protein VPI. We find that both these cases, as well as one of the original cases reported in 1972, show evidence of the presence of JCV but not SV40. Thus, one of the original cases was misidentified, and the single remaining case, consisting of biopsy material no longer available, is without confirmation. No cases of PML in AIDS which we or others have studied have been identified as PML-SV40. There are now no credible reports of PML due to SV40 in man. If JCV interacts uniquely with the brain in PML, as determined here, it becomes more likely that JCV also makes a unique contribution in other diseases (e.g., multiple sclerosis, medulloblastoma) in which the virus may remain in a latent state without overt viral replication. It is therefore proposed that several diseases may result from differing host responses to the same virus focally distributed in the brain following infection of glial cells or their precursors early in life. While the host responses maybe crucial, it is also true that different virus strains may have different degrees of neurotropism and different expression within the human brain. That aspect of virus-host interaction is now under investigation.