In previous studies, we have shown that a herpesvirus vector can transfer a therapeutic cellular gene (beta-glucuronidase) from peripheral sites of inoculation into the central nervous system in mice with a model neurodegenerative disease caused by a deficiency of this enzyme (mucopolysaccharidosis type VII, Sly disease). The vector corrects the enzymatic deficiency in transduced cells but the number of cells corrected is too low to alter the pathology of the disease. The recombinant vector virus, which has the foreign gene substituted into the viral LAT locus, had reduced pathogenicity after corneal inoculation compared to the wild-type virus from which it was derived (HSV-1 strain 17+). We therefore attempted to increase the number of corrected cells in the MPS VII brain by increasing the inoculating dose of the vector. However, the vector was acutely pathogenic in the diseased mice at doses that were non-pathogenic in normal littermates. The pathogenic effect of the vector virus in the mutants could be blocked by passive immunization with human gamma-globulin containing anti-HSV-1 antibodies on the day of infection but not when given at the peak of viral replication (day 4). However, effective protection also blocked transduction by the vector, thereby abrogating the effects of increased vector dosage. The effect was virus specific because inoculation of a high dose of a non-pathogenic variant of strain 17+ virus (1716) directly into the brains of MPS VII mice was not lethal. We found no apparent differences in the acute inflammatory response in mutant versus normal animals. These data suggest that the increased susceptibility to vector virulence was related to the overall compromised state of health of the diseased animals, which is further supported by the observations that the mutant mice are more sensitive to stress and to anesthetics than normal littermates. These findings indicate that adverse effects of gene transfer vectors for genetic diseases may not be fully apparent when tested in normal animals.