Epidemiological studies of the uncommon disorder neuromyelitis optica spectrum disorder (NMOSD) may be difficult to interpret because of the evolving nature of diagnostic criteria, differences in the definition and accuracy of NMOSD diagnosis, the completeness of case ascertainment, and variability in assays for the disease‐specific biomarker aquaporin‐4 (AQP4)‐IgG. A sub‐group of patients with the clinical syndrome NMOSD lack detectable AQP4‐IgG and in these cases an accurate diagnosis requires precise diagnostic algorithms and longitudinal follow‐up. Consecutive sets of criteria for NMO/NMOSD have been introduced during the two last decades. Such criteria need validation in different populations. Detection of other autoantibodies, such as IgG specific for myelin oligodendrocyte glycoprotein or for glial fibrillary acidic protein in a sub‐group of AQP4‐IgG–negative NMOSD patients, has improved over the past decade and may lead to overlap of the clinical syndromes/phenotypes. This review begins by summarizing current knowledge on the widening clinical spectrum of NMOSD. Subsequently, we describe two epidemiological studies from Denmark carried out in two different decades (1998–2008 and 2007–2014) and comment on the differences in study design, patient ascertainment, and interpretation of results. These factors may explain some of the observed differences, reflecting the complexity and providing a clear example of this development.