Background Neuromyelitis optica is an inflammatory demyelinating disease with generally poor prognosis that selectively targets optic nerves and spinal cord. It is commonly misdiagnosed as multiple sclerosis. Neither disease has a distin guishing biomarker, but optimum treatments differ. The relation of neuromyelitis optica to optic spinal multiple sclerosis in Asia is uncertain. We assessed t he capacity of a putative marker for neuromyelitis optica (NMO IgG) to disting uish neuromyelitis optica and related disorders from multiple sclerosis. Methods Indirect immunofluorescence with a composite substrate of mouse tissues identif ied a distinctive NMO IgG staining pattern, which we characterised further by dual immunostaining. We tested masked serum samples from 102 North American patients with neuromyelitis optica or with syndromes that suggest high risk o f the disorder, and 12 Japanese patients with optic spinal multiple sclerosis. Control patients had multiple sclerosis, other myelopathies, optic neuropathies , and miscellaneous disorders. We also established clinical diagnoses for 14 pat ients incidentally shown to have NMO IgG among 85000 tested for suspected para neoplastic autoimmunity. Findings NMO IgG outlines CNS microvessels, pia, subp ia, and Virchow Robin space. It partly colocalises with laminin. Sensitivity a nd specificity were 73% (95% CI 60- 86) and 91% (79- 100) for neuromyeli tis optica and 58% (30- 86)- and 100% (66- 100) for optic spinal multip le sclerosis.NMO IgG was detected in half of patients with high risk syndrom es. Of 14 seropositive cases identified incidentally,12 had neuromyelitis optica or a high risk syndrome for the disease. Interpretation NMO IgG is a specif ic marker autoantibody of neuromyelitis optica and binds at or near the blood brain barrier. It distinguishes neuromyelitis optica from multiple sclerosis. As ian optic spinal multiple sclerosis seems to be the same as neuromyelitis opti ca. Background Neuromyelitis optica is an inflammatory demyelinating disease with generally poor prognosis that selectively targets optic nerves and spinal cord. It is commonly misdiagnosed as multiple sclerosis. Neither disease has a distin guishing biomarker, but optimum law differ. The relation of neuromyelitis optica to optic  Spinal multiple sclerosis in Asia is uncertain. We assessed t he capacity of a putative marker for neuromyelitis optica (NMO-IgG) to disting uish neuromyelitis optica and related disorders from multiple sclerosis. Methods Indirect immunofluorescence with a composite substrate of mouse tissues identified a distinctive NMO IgG staining pattern, which we characterized further by dual immunostaining. We tested masked serum samples from 102 North American patients with neuromyelitis optica or with syndromes that suggest high risk of the disorder, and 12 Japanese patients with optic-spinal multiple sclerosis. Control patients had multiple sclerosis, Other myelopathies, optic neuropathies, and miscellaneous disorders. We also established clinical diagnoses for 14 pat ients incidentally shown to have NMO-IgG among 85,000 tested for suspected para neoplastic autoimmunity. Findings NMO-IgG outlines CNS microvessels, pia, subp ia, and Virchow Sensitivity a nd specificity were 73% (95% CI 60-86) and 91% (79-100) for neuromyelis optica and 58% (30-86) - and 100% ( 66- 100) for optic-spinal multip le sclerosis. NMO IgG was detected in half of patients with high-risk syndrom es. Of 14 seropositive cases identified incidentally, 12 had neuromyelitis optica or a high-risk syndrome for the disease. Interpretation It distinguishes neuromyelitis optica from multiple sclerosis. As ian optic-spinal multiple sclerosis seems to be the same as neuromyelitis opti ca.