Of the hypotheses put forward to explain why occasional individuals with two X chromosomes are nonetheless male, the one that has attracted most attention is the possibility that one of the X chromosomes has obtained a small piece of Y chromosome which is sufficient to produce 'maleness'. This hypothesis was based primarily on the observation that in two families with XX males both fathers were Xg(a+) and both probands Xg(a-). (Xg shows X-linked dominant inheritance.) This theory holds that an anomalous X-Y interchange at meiosis in the father resulted in the paternal X chromosome's losing the Xg gene and acquiring a male-determining gene from the Y chromosome. While, for example, the frequencies of Xg phenotypes among XX males and the cytogenetic observation of a structural abnormality in one X are compatible with this hypothesis, direct evidence of it is lacking. Here we describe an XX male who expresses his father's allele for 12E7, a Y-linked marker, but fails to express his father's allele for Xg, an X-linked marker. These findings strongly suggest that anomalous X-Y interchange occurred in this case and perhaps in that of many other XX males. We suggest that a male-determining gene on the Y has also been translocated to the X and caused maleness in the proband. These results are discussed in the light of current models of X-Y chromosomal homology.