(Extracted from《SCIENCE》VOL. 10, NO. 466, by Rebecca M. Jones) Behavioral interventions effectively reduce severity of autism symptoms and improve social functioning. Music intervention has been shown to improve social interactions in children with autism, and neuroimaging studies in healthy controls have shown that circuits important for emotion and memory processing are engaged during musiclistening. Yet, it is unknown whether a music intervention can alter brain activity in children with autism. In a randomized controlled trial, Sharda et al. evaluated the effects of 8- to 12-week music intervention in 6- to 12-year-old children with autism, comparing music therapy to a nonmusic control intervention to determine if music intervention would improve social communication skills and alter brain connectivity. Before and after the intervention, caregivers reported about their child’s behavior, and children completed a resting state MRI scan. The music intervention used musical instruments, songs, and rhythmic cues whereas the nonmusic control wasa series of play-based activities that was designed to be as structurally matched as possible to the music intervention. Children who were in the music intervention group (n = 24) had improvements in parent-reported language, social relationships, and family quality of life compared with the control group (n = 23). Functional connectivity was greater between auditory and subcortical regions and auditory and motor regions in the music versus nonmusic group and lower between auditory and visual areas, circuitry that isover-connected in autism. Greater connectivity between auditory and subcortical areas was associated with greater improvements in social communication scores. Thus, in a relatively short period of time, music intervention changed social communication behavior and altered circuitry important for processing rewards. These findings have significant implications for using music as a therapeutic tool in autism. A significant limitation of this work is that the primary behavioral outcome measure was parent-reported, which can be biased by parental expectations. Future work testing how long the behavior and brain changes are sustained will be critical for determining the efficacy of music intervention in autism.