Previous studies of the gap-junction beta-2 subunit gene GJB2 (connexin 26) have suggested that the 101T-->C (M34T) nucleotide substitution may be a mutant allele responsible for recessive deafness DFNB1. This hypothesis was consistent with observations of negligible intercellular coupling and gap-junction assembly of the M34T allele product expressed in Xenopus oocytes and HeLa cells. The results of our current study of a family cosegregating the 167delT allele of GJB2 and severe DFNB1 deafness demonstrate that this phenotype did not cosegregate with the compound-heterozygous genotype M34T/167delT. Since 167delT is a null allele of GJB2, this result indicates that the in vivo activity of a single M34T allele is not sufficiently reduced to cause the typical deafness phenotype associated with DFNB1. This observation raises the possibility that other GJB2 missense substitutions may not be recessive mutations that cause severe deafness and emphasizes the importance of observing cosegregation with deafness in large families to confirm that these missense alleles are mutant DFNB1 alleles.