ObjectiveTo assess video-polysomnographic (vPSG) criteria and their cutoff values for the diagnosis of disorders of arousal (DOAs; sleepwalking, sleep terror). MethodsOne hundred sixty adult patients with DOAs and 50 sex- and age-matched healthy participants underwent a clinical evaluation and vPSG assessment to quantify slow wave sleep (SWS) interruptions (SWS fragmentation index, slow/mixed and fast arousal ratios, and indexes per hour) and the associated behaviors. First, a case-control analysis was performed in 100 patients and the 50 controls to define the optimal cutoff values using receiver operating characteristic curves. Their sensitivity was then assessed in the other 60 patients with DOAs. ResultsThe SWS fragmentation index and the mixed, slow, and slow/mixed arousal indexes and ratios were higher in patients with DOAs than controls. The highest area under the curve (AUC) values were obtained for the SWS fragmentation and slow/mixed arousal indexes (AUC = 0.88 and 0.90, respectively). The SWS fragmentation index cutoff value of 6.8/h reached a sensitivity of 79% and a specificity of 82%. The slow/mixed arousal index had a sensitivity of 94% for the 2.5/h cutoff, and 100% specificity for 6/h. Both parameters showed good interrater agreement, and their sensitivities were confirmed in the second group of patients. Combining electroencephalographic parameters and video-based behavioral analyses increased the correct classification rate up to 91.3%. InterpretationFrequent slow/mixed arousals in SWS and complex behaviors during vPSG are strongly associated with DOAs, and could be promising biomarkers for the diagnosis of non-rapid eye movement parasomnias. Ann Neurol 2018;83:341-351