SummaryThe congenital deaf mute (HD) population of Northern Ireland, reported by Stevenson & Cheeseman, appears by maximum likelihood segregation analysis and the theory of detrimental equivalents to consist of three types.1. Autosomal recessives, complete penetrance, 68% of all HD cases, at least several loci, average mutation rate 1 × 10‐5 per locus, carrier frequency 0–080 per gamete.2. Autosomal dominants, high but not complete penetrance, 22 % of all HD cases, at least two loci, total mutation rate 5 × 10_5 per gamete.(a) Genes inherited from affected parents, 7 % of all cases.(b) New mutants inherited from normal parents, 15 % of all cases.3. Sporadic cases, due to unrecognized infection or more complex genetic mechanisms, 9 % of all cases.On the evidence, the hypothesis that the (non‐significant) excess of affected boys is due to sex‐linkage can neither be proven nor excluded. Ex hypothesi, sex‐linked cases account for less than 2 % of all HD cases, with a mutation rate of 2 × 10‐6 per X chromosome per generation.The hypotheses of heterozygote advantage, absence of dominant cases, synergistic effect of nonallelic recessive genes, and hypotruncate selection are shown to be inconsistent with the data. Studies on deaf mutism in other populations do not disagree with these conclusions.The theory of detrimental equivalents is extended to rare traits with retrospective ascertainment of consanguinity, giving estimates of carrier frequencies, number of loci, and mutation rates for recessive genes.