The last 25 years have witnessed tremendous changes in our ability to detect autism very early in life and provide interventions that can significantly influence children's outcomes. It was once questioned whether autism could be recognized before children had developed language and symbolic play skills; now changes in early behaviors, as well as structural brain changes, have been documented in infants 6-12 months of age who later develop autism. Advances in brain imaging and genetics offer the possibility of detecting autism before the syndrome is fully manifest, thereby reducing or preventing symptoms from developing. Whereas the primary mode of behavioral intervention a few decades ago relied on operant conditioning, recent approaches integrate the methods of applied behavioral analysis within a developmental, relationship-focused intervention model that are implemented by both parents and clinicians. These interventions have been found to have positive effects on children's developmental trajectory, as measured by both behavioral and neurophysiological assessments. Future approaches will likely combine both behavioral and pharmacological treatments for children who have less robust responses to behavioral interventions. There has been a paradigm shift in the way that autism is viewed, evolving from a lifelong condition with a very poor prognosis to one in which significant gains and neuroplasticity is expected, especially when the condition is detected early and appropriate interventions are provided. The grand challenge for the future is to bridge the tremendous gap between research and the implementation of evidence-based practices in the broader community, both in the United States and worldwide. Significant disparities in access to appropriate health care for children with autism exist that urgently require advocacy and more resources.