Pro-inflammatory T cells in the central nervous system (CNS) are causally associated with multiple demyelinating and neurodegenerative diseases. Accumulating evidence has determined the effector pathways by which T cells drive neuroinflammation and revealed unexpected roles for CNS-extrinsic factors, such as diet and the microbiota, in impacting these T cell responses. Here we explored the contribution of a recently appreciated innate immune cell type on T cell-mediated neuroinflammation, termed innate lymphoid cells (ILCs). ILCs are enriched at mucosal barrier surfaces of the mammalian body, are predominantly tissue resident, exhibit comparable heterogeneity to CD4 T cell effector cells, and have substantial bidirectional communication with adaptive immunity. We defined a unique population of group 3 innate lymphoid cells (ILC3s) that infiltrate the CNS in a mouse model of multiple sclerosis (MS). These ILC3s are derived from the circulation, localize in proximity to infiltrating T cells in the CNS, function as antigen presenting cells that restimulate myelin-specific T cells, and are elevated in the blood and cerebral spinal fluid of individuals with multiple sclerosis. Critically, antigen presentation by ILC3s is required to promote pro-inflammatory T cell responses in the CNS and the development of multiple sclerosis-like disease in mouse models. In contrast, conventional and tissue-resident ILC3s in the periphery do not appear to contribute to disease induction, but instead limit autoimmune T cell responses and prevent multiple sclerosis-like disease when experimentally targeted to present myelin antigen. Collectively, we have identified two major functional roles for antigen-presenting ILC3 during neuroinflammation that occur in a context and cellular subset specific manner. This includes an inflammatory ILC3 population that infiltrates the CNS and is essential for directly promoting T cell-dependent neuroinflammation, and conversely, a peripheral tissue-resident ILC3 population that has the potential to prevent autoimmune disease if targeted with antigen. Our expanded model on the function of ILC3 subsets in the context of autoimmune neuroinflammation provides both a greater understanding of, and potential therapeutic opportunities for, MS and other demyelinating or neurodegenerative diseases in which self-specific T cell responses contribute to disease pathogenesis.