Animal models of neuromyelitis optica (NMO) are needed for elucidation of disease mechanisms and for testing new therapeutics. Prior animal models of NMO involved administration of human anti-aquaporin-4 immunoglobulin G antibody (NMO-IgG) to rats with pre-existing neuroinflammation, or to naïve mice supplemented with human complement. We report here the development of NMO pathology following passive transfer of NMO-IgG to naïve rats. A single intracerebral infusion of NMO-IgG to adult Lewis rats produced robust lesions around the needle track in 100 % of rats; at 5 days there was marked loss of aquaporin-4 (AQP4), glial fibrillary acidic protein (GFAP) and myelin, granulocyte and macrophage infiltration, vasculocentric complement deposition, blood–brain barrier disruption, microglial activation and neuron death. Remarkably, a distinct ‘penumbra’ was seen around lesions, with loss of AQP4 but not of GFAP or myelin. No lesions or penumbra were seen in rats receiving control IgG. The size of the main lesion with loss of myelin was greatly reduced in rats made complement-deficient by cobra venom factor or administered NMO-IgG lacking complement-dependent cytotoxicity (CDC) effector function. However, the penumbra was seen under these conditions, suggesting a complement-independent pathogenesis mechanism. The penumbra was absent with NMO-IgG lacking both CDC and antibody-dependent cellular cytotoxicity (ADCC) effector functions. Finally, lesion size was significantly reduced after macrophage depletion with clodronate liposomes. These results: (i) establish a robust, passive-transfer model of NMO in rats that does not require pre-existing neuroinflammation or complement administration; (ii) implicate ADCC as responsible for a unique type of pathology also seen in human NMO; and (iii) support a pathogenic role of macrophages in NMO.