Dyskeratosis congenita (DC) is a rare inherited bone marrow failure (BMF) syndrome. The classical features of DC include nail dystrophy, abnormal skin pigmentation, and mucosal leukoplakia. The diagnosis of DC can be difficult. Originally, the diagnosis was based on the presence of the classical mucocutaneous features. However, the identification of four genes responsible for DC (DKC1, TERC, TERT, and NOP10) showed that these mucocutaneous features are only present in a proportion of patients with DC. Additionally, screening for mutations in the affected genes is expensive and is negative in about 50% of patients with classical features of DC. The products of the genes mutated in DC are the components of the telomerase ribonucleoprotein complex, which is essential for telomere maintenance. Therefore it has been postulated that DC is a disease arising from excessive telomere shortening. Here we examined whether the measurement of telomeres could be used as a screening test to identify individuals with DC. For this purpose we examined telomere length in peripheral blood mononuclear cells from 169 patients who presented with bone marrow failure including 17 patients with DC, diagnosed by the presence of classical cutaneous features or the identification of mutations in DKC1, TERC or TERT, 28 patients with paroxysmal nocturnal hemoglobinuria, 25 patients with Diamond Blackfan anemia, 5 patients with Shwachman-Diamond syndrome, 8 patients with myelodysplastic syndrome, and 74 patients with aplastic anemia of unknown cause classified as idiopathic aplastic anemia. In addition we measured telomere length in 12 patients with idiopathic pulmonary fibrosis and in 45 individuals with a de novo deletion of chromosome 5p including the TERT gene. Their telomere lengths were compared with those of 202 age-matched healthy controls. Moreover, mutations were screened in the genes associated with DC. In cases where a mutation was identified, telomere length and mutations were also examined in all the family members. Our results show that all patients with DC and bone marrow failure have very short telomeres far below the first percentile of healthy controls. Not all mutation carriers, including some carriers of apparently dominant mutations, have very short telomeres. What is more, very short telomeres could be found in healthy individuals in these families, some of whom were not mutation carriers. These findings indicate that in patients with BMF the measurement of telomere length is a sensitive screening method for DC, whether very short telomeres in this setting are also specific for DC remains to be determined. However, in contrast to a previous study, we find that telomere length does not always identify mutation carriers in the families of DC.