Autism is a developmental disorder, but the least is known about what is most important: Development. What are the early brain abnormalities? What neural functions do they disrupt? What the first behavioral indicators of risk for autism? What is the prognosis for the toddler or child at first diagnosis? Which are likely to respond to known effective treatment and which not? Are their brain or other biological markers that could predict non-responders so that treatment research could be targeted towards discovery of treatments that could help them? What genes and gene pathways are responsible for early brain defects? The answer to each question comes to the same point: Early identification methods. These sorts of crucial questions cannot be addressed unless there is a way to identify infants at risk for autism. The most popular method is the infant sibling method. Now many groups are trying this method. The first two laboratories to use this method began their research five and seven years ago, and have published two original papers, one on 7 autism spectrum (ASD) infants and one on 27 ASD infants. They found no differences from typical development at 6 months, but did by 12 months. They provided no brain, neurofunctional, genetic or other biological data on the ASD infants. The infant sibling method has major cost, practicality and methodological limitations, which are detailed in our grant. Although popular and in progress for a long time, to date, studies using this method have provided no answer any of the major questions posed above. A simpler, quicker, more flexible, cost-effective and more clinically relevant and practical method is to study individuals referred by physicians because they display behavioral symptoms indicating risk for an ASD. In the 1990s, this method resulted in identification of mostly at-risk 3 year olds and up. By the early 2000s, the at-risk age for referral was age 2 years because of heightened awareness by physicians and new diagnostic tools. In each "era", others and we were able to vigorously investigate brain and behavior abnormalities and new treatment approaches at young ages in ASD. It was via this simple referral method that major discoveries about early brain overgrowth in ASD occurred in our laboratory. It was via this method that we have been able to perform the first fMRI activation studies on ASD at age 2-years. Now, via a novel variant of this proven effective prospective method (see Core B), in our ACE Center we will study 12-month old infants at-risk for ASD, infants at-risk for developmental delay and typical infants using advanced biological and behavioral methods. Each infant will be studied at intervals longitudinally and a final best estimate diagnosis made at 36 months providing a final step for group assignment for statistical