Recent MRI studies suggest that the brain of a toddler with autism grows at an excessive rate, leaving the child with an enlarged brain volume relative to typically-developing toddlers. The frontal and temporal lobes are sites of maximal brain enlargement. The frontal lobe plays vital roles in higher-order cognitive, language, social, and emotional functions, each of which is seriously deficient in autistic subjects. One hypothesis to account for the enlargement is that the autistic brain has an excessive number of neurons. However, following the overgrowth phase, there is also evidence of a premature growth arrest in frontal lobes. Thus, there is the intriguing possibility that while there are an excessive number of neurons in autism during early childhood in frontal cortex, this phase is followed by a decline in numbers beginning by preadolescence. The proposed study will test this hypothesis by carrying out a systematic stereological examination of the number of neurons in the entire frontal cortex as well as five well-defined frontal regions in 16 autistic and 16 control cases ages 2 years to 45 years, examining both group differences and age related changes in neuron number. In addition, since autism cases included in the study will have met ADI-R criteria, the investigators will carry out biostatistical analyses of the relationship between neuron numbers, clinical phenotypic information (from the ADI-R), and postmortem variables of interest (e.g., cause of death, post-mortem interval). What this means for people with autism: The neural defects driving early brain overgrowth may underlie the behavioral and clinical emergence of autism. This study will begin to reveal the underlying developmental cellular and molecular defects that can contribute to the development of autistic behavior.