The development of rapid procedures for the characterization of mutations is advancing the knowledge of the molecular biology of the haemophilias and transforming the strategies for the diagnoses required for genetic counselling. In haemophilia B more than 300 mutants have been fully characterized. These comprise complete and partial deletions, rare insertions, and 'point' mutations. The latter may impair transcription (promoter mutations), RNA processing (splicing mutations) and translation (frameshifts and stop codons) or cause single amino acid (aa) changes. Eighty-four residues are involved in the 105 presumed detrimental aa substitutions reported so far and these are usually conserved in the factor IX homologues (factors VII, X and protein C) and/or the factor IX of different mammalian species. There are clear correlations between the mutation and clinical features. In addition mutations causing gross physical or functional loss of coding information appear to predispose to the development of antibodies against therapeutic factor IX. Hotspots of mutations have been identified and are usually associated with CpG sequences. In haemophilia A the size and complexity of the factor VIII gene has hindered the analysis of mutants. Most of the studies published so far have analysed only a small fraction of the essential region of the factor VIII gene and this led to the repeated observation of specific types of mutation. The recent development of a rapid method to analyse RNA splicing and the whole coding region of the factor VIII gene should unblock this situation. With regard to genetic counselling, the direct detection of gene defects has increased the proportion of haemophilia B families that can be helped from 60% to virtually 100% and similar expectations may now be formulated for haemophilia A. In the UK a national database of haemophilia B mutations is being constructed to optimize genetic counselling. This should offer a model for a similar development in haemophilia A.