This project will explore ways to more accurately determine environmental causes of autism using data from the Early Autism Risk Longitudinal Investigation (EARLI), a study of pregnant mothers who already have a child with autism spectrum disorders (ASD) and studying the environmental toxicants which they have been exposed to. Studies of this type are very difficult to conduct because even small errors in measurement of the potential toxin, the type of exposure or when it occurs can make a big difference in observed associations between the exposure and autism. Further complicating these studies, biological samples are taken from mothers and children in small amounts, making multiple tests difficult. The purpose of this project is twofold. First, we will examine how sensitive observed associations may be to small changes in measurement, with the goal of choosing the measurement strategies with the least amount of error. Second, we will develop strategies to test for the effects of multiple environmental exposures using the same biological sample, so as to conserve this finite resource. By conducting this type of methodological work early in the study, we will make much better use of limited resources and have much more confidence in our results. This will also provide guideline for others to study the environmental impacts on the development of autism.