Consanguineous marriages, usually between first cousins or between uncle and niece, are common in certain societies. The transmission/disequilibrium test (TDT) compares the transmission from parents to an affected child of alleles at a marker locus, and differential transmission indicates linkage and allelic association between the marker locus and a disease locus. We investigate the consequences for the TDT, as a test for allelic association, of consanguineous marriages. For each parental marker mating type, we calculate the frequency of each disease mating type, and the associated probability that an offspring is affected. We use Bayes' Theorem to estimate the probability that an affected child inherits the given allele from a heterozygous parent, then combine our results across marker mating types. The effect of consanguineous marriage is usually small. For candidate genes, the effects were greater for uncle-niece marriages, for rare disease alleles, and for high genotype relative risk. For markers, the effects were generally negligible. The Type I error probability of the TDT is essentially unchanged by intermarriage, except for a purely recessive disease allele. The power of the TDT is increased for a recessive allele and decreased for a dominant allele. However, consideration of levels of consanguinity that arise in practice indicates that standard power calculations for the TDT will usually need only minor modification.