Multiple sclerosis is the commonest neurological disease affecting young adults. Whilst the initial relapsing-remitting disease phase is associated with inflammatory demyelination and is treatable with immunomodulatory drugs, the secondary progressive phase (SP-MS) is associated with ongoing axonal loss and cortical atrophy and is currently untreatable. Studies of SP-MS have revealed the presence of extensive subpial demyelinated lesions within the cerebral cortex. This pathology is associated with a high level of meningeal inflammation, a gradient of cell loss from the cortical surface and high levels of microglia activation. To test the hypothesis that pro-inflammatory cytokines diffusing from the cerebral meninges could be responsible, we have established an animal model mimicking cortical grey matter pathology. Female DA rats were immunised with 5μg recombinant myelin oligodendrocyte glycoprotein (rmMOG) in incomplete Freunds adjuvant (IFA). This dose was insufficient to initiate encephalomyelitis, but did initiate an anti-MOG humoral immune response in the periphery. Twenty days post-immunisation animals received an injection of tumour necrosis factor (TNF) and interferon gamma (IFNγ) into the subarachnoid space at the sagittal sulcus. Immunohistochemistry revealed areas of subpial demyelination extending through cortical layers I–III. Lesions were maximal after 7 days and had resolved by remyelination at 14 days. A gradient of microglia/macrophage activation was present from the cortical surface. The extent of demyelination correlated with activation of microglia in the cortex and macrophages within the meninges. Activated microglia were observed contacting myelin, oligodendrocytes and neurons. In the demyelinated cortex, expression of the TNF receptors TNFR1A and TNFR1B was upregulated on oligodendrocytes and perivascular macrophages respectively. CD8+ T cells were observed in the meninges, corpus callosum and scattered throughout the grey matter, whereas CD4+ T cells and CD79a+ B cells were restricted to the meninges. Oligodendrocyte numbers were reduced in the upper cortical layers prior to demyelination (days 1 and 3 post-injection), but were still present in demyelinated lesions at day 7. Numbers of neurons and astrocytes were not changed. Control animals immunised with IFA and injected with cytokines had increased presence of inflammatory cells within the meninges but no demyelination. Animals immunised with rmMOG and injected with PBS had no demyelination or immune response within the meninges or cortex. Thus, acute subpial demyelination was dependent on a pre-existing immune response against myelin protein, coupled with generalised pro-inflammatory signalling within the meninges. These findings support our hypothesis of a role for meningeal inflammation in the cortical pathology of MS and describe for the first time an animal model that can be used to study the molecular mechanisms involved. Future research will aim to maintain meningeal inflammation and produce a model of chronic demyelination.