Fanconi anaemia (FA) is a rare autosomal recessive or X-linked inherited disease characterised by an increased incidence of bone marrow failure (BMF), haematological malignancies and solid tumours. Cells from individuals with FA show a pronounced sensitivity to DNA interstrand crosslink (ICL)-inducing agents, which manifests as G2-M arrest, chromosomal aberrations and reduced cellular survival. To date, mutations in at least 15 different genes have been identified that cause FA; the products of all of these genes are thought to function together in the FA pathway, which is essential for ICL repair. Rapidly following the discovery of FA genes, mutant mice were generated to study the disease and the affected pathway. These mutant mice all show the characteristic cellular ICL-inducing agent sensitivity, but only partially recapitulate the developmental abnormalities, anaemia and cancer predisposition seen in individuals with FA. Therefore, the usefulness of modelling FA in mice has been questioned. In this Review, we argue that such scepticism is unjustified. We outline that haematopoietic defects and cancer predisposition are manifestations of FA gene defects in mice, albeit only in certain genetic backgrounds and under certain conditions. Most importantly, recent work has shown that developmental defects in FA mice also arise with concomitant inactivation of acetaldehyde metabolism, giving a strong clue about the nature of the endogenous lesion that must be repaired by the functional FA pathway. This body of work provides an excellent example of a paradox in FA research: that the dissimilarity, rather than the similarity, between mice and humans can provide insight into human disease. We expect that further study of mouse models of FA will help to uncover the mechanistic background of FA, ultimately leading to better treatment options for the disease.