What causes autism? This question has been a topic of debate that has been stirring in the media for the past few years. What is clear is that parents usually notice signs of autism in the first few years of life, and clinicians diagnose autism based on behavioral criteria such as repetitive movements, restricted interests, and poor social interactions. Scientists have discovered that autism has a genetic basis, as children with autism seem to be either missing- or have extra copies of certain portions of their chromosomes. One of the best examples, discovered by Mike Wigler's group at Cold Spring Harbor, is deletion of a portion of human chromosome 16 that includes the genes that are blue prints for 28 different proteins. Is this just a coincidence, or does having too few of these 28 proteins really cause autism? We directly asked this question by creating mice that were missing the same 28 genes. Remarkably, these mice have unique behaviors and unusual brain architectures that demonstrate that deletion of this 28-gene region causes autism-like features. It is possible that one of these 28 genes is responsible for autism. Another possibility is that several of these genes, when missing, lead to autism. We will directly test these possibilities by creating mice that are missing only some of the 28 genes, and asking if they behave the same and have the same brain alterations as the mice that are missing the entire 28-gene region. These studies will allow us to identify the single gene--or alternatively the group of genes--that when missing causes autism. Once we know this, it is possible to design better ways to diagnose and treat children with autism.