Two of the genes responsible for the bone marrow failure syndrome dyskeratosis congenita (DC) have been identified. These have evoked considerable interest as the products of both of these genes are involved in the telomerase complex. Telomerase is responsible for maintaining telomere length at the ends of each chromosome and hence the replicative potential of stem cells and cells in the germ line. In order to define the heterogeneity of DC, we have reviewed the clinical features and genetic basis of 213 families that have been referred to the DC registry at the Hammersmith Hospital. In 67 families we have identified 34 different mutations in the gene that encodes the protein dyskerin (DKC1), 9 of which are unpublished (P10L, T67I, H68Q, K314R, D359N, A386T, T408I, S420Y and IVS14nt473A->G). While most of these mutations are unique, the remarkably frequent A353V substitution is seen in 28 different families. Presentation of cases with this amino acid substitution ranges from a very severe disease, causing death at the age of 4 years, to a man who is alive into his 40s. In 10 families we have found mutations in the gene encoding the RNA component of telomerase (TERC), all of which are unique. Presentation of these heterozygous TERC mutations is even more variable, ranging from an asymptomatic 80 year old woman, to a girl presenting with hypoplastic MDS at the age of 3. In a substantial proportion of families (40/213) the index case presents with very severe disease, previously defined as the Hoyeraal-Hreidersson (HH) syndrome (characterized by growth retardation, microcephaly, cerebellar hypoplasia, and the development of bone marrow failure/immunodeficiency usually below the age of 5 yrs). In 28 of these families there is affected male(s) only, 13 of which have DKC1 mutations. Of these, 4 have the same amino acid substitution (T49M), which is only seen in patients with HH. In the 12 HH families with affected females, none have DKC1 mutations, 6 are sporadic and 6 appear to have an autosomal recessive form of inheritance. In a large number of families we have been unable to find either a DKC1 or TERC mutation. This includes not only the majority (51/60) of the families in which there is an affected female, but also nearly two thirds (77/113) of the sporadic male cases. Other important gene(s), possibly related to the telomerase complex, are therefore yet to be identified.