BACKGROUND: The threat of emerging infections and recognition of novel immune-mediated forms of encephalitis has raised the profile of this condition in recent years. Incidence is poorly defined and most cases have an unknown cause. There is currently much interest in identification of new microbial agents of encephalitis, but no work has investigated systematically reasons for lack of pathogen identification in studies. METHODS: We systematically reviewed published literature on incidence and etiology of encephalitis in non-outbreak settings and explored possible explanations for the large number of cases of unknown etiology. RESULTS: Annual incidence ranged from 0.07 to 12.6 cases per 100,000 population with an evident decrease over time (p = 0.01). The proportion of cases with unknown etiology was high across studies (>50% in 26 of 41 studies), with strong evidence of heterogeneity in study findings (p < 0.001). Our meta-regression identified study period, setting, and subsyndrome to be the main contributors to between-study variation, rather than methodologic factors such as study design, case definitions, sample types, and testing strategies. CONCLUSIONS: Our findings support the hypothesis that new and emerging infectious agents, or new forms of immune-mediated encephalitis, may be responsible for cases currently of unknown cause and encourage the ongoing global effort to identify these. Our review highlights research areas that might lead to a better understanding of the causes of encephalitis and ultimately reduce the morbidity and mortality associated with this devastating condition.