A pseudoautosomal locus for schizophrenia has been proposed based on observations of an excess of same-sex affected sibling pairs over opposite-sex pairs when the transmitting parent is the father. Such a pattern of partial concordance by sex related to paternal transmission would be difficult to explain by any biologic mechanism other than pseudoautosomal transmission of schizophrenia. To test the pseudoautosomal hypothesis, 37 sibling pairs concordant for the schizophrenia spectrum were identified from 24 nuclear pedigrees. No significant difference in concordance for sex was found between sibships of paternal and maternal transmission of schizophrenia. Next, a linkage analysis was performed in 12 informative pedigrees, examining seven marker loci spanning the pseudoautosomal region. Both strict schizophrenia and a broader schizophrenia spectrum were analyzed as the affected phenotype, and both autosomal dominant and autosomal recessive models were tested. None of the markers supported linkage to either schizophrenia or the schizophrenia spectrum. Lod scores of less than -4 were obtained across the entire pseudoautosomal region by means of multipoint linkage analyses in the autosomal dominant model. In the autosomal recessive model, the respective lod scores were less than -2. We conclude that there is no evidence of a pseudoautosomal locus for schizophrenia in our pedigrees in any of the genetic models we tested.