Dexmedetomidine has been increasingly in use for pediatric noninvasive procedural sedation. This retrospective study examined experience in children with autism and other neurobehavioral disorders, populations often difficult to sedate. Records of children with autism or neurobehavioral disorders sedated with dexmedetomidine at Chris Evert Children's Hospital and Kosair Children's Hospital were reviewed. Demographic and sedation-related data were collected, including sedative doses, time to sedation, efficacy, and complications. Comparisons of sedative doses, efficacy between autism and neurobehavioral patients, and analysis of age-related factors were performed. In all, 315 patients were sedated, most commonly for magnetic resonance imaging. Mean induction and total dexmedetomidine doses were 1.4 +/- 0.6 and 2.6 +/- 1.6 microg/kg, respectively, with no differences between autism and neurobehavior patients. Most patients (90%) patients received concomitant midazolam. There was an age-related decrease in dexmedetomidine dose, independent of midazolam use. Seven patients required intervention for hypotension, bradycardia, or both, and only one adverse respiratory event (obstruction requiring nasopharyngeal airway placement) occurred. There were two episodes of overt recovery-related agitation. All but four procedures were successfully completed (4/315, or 98.7%). Dexmedetomidine with or without midazolam was an effective sedative in this population. The regimen appeared to be well tolerated with few adverse events, including recovery-related agitation, and appears to be an attractive option for this population.