Psychometric tests are routinely used to assess facets of neurodevelopment in epidemiological studies and are a tool for estimating the potential effects of toxicants on the nervous system. When assessing the validity and reliability of results from a series of epidemiological studies, the specific psychometric tests utilized and factors affecting their administration in human populations present unique challenges. This report describes the historical application of psychometric tests to the study of the neurodevelopmental toxicity of methylmercury and defines neurodevelopmental domains that are assessed with these instruments. Principles are proposed for evaluating the validity and reliability of psychometric tests and for identifying potential sources of bias in the selection and administration of these tests in human populations.