In this article we examine sampling strategies and analytical methods used in a series of recent studies of children's exposure to pesticides that may prove useful in the design and implementation of the National Children's Study. We focus primarily on the experiences of four of the National Institute of Environmental Health Sciences/U.S. Environmental Protection Agency/ Children's Centers and include University of Washington studies that predated these centers. These studies have measured maternal exposures, perinatal exposures, infant and toddler exposures, and exposure among young children through biologic monitoring, personal sampling, and environmental monitoring. Biologic monitoring appears to be the best available method for assessment of children's exposure to pesticides, with some limitations. It is likely that a combination of biomarkers, environmental measurements, and questionnaires will be needed after careful consideration of the specific hypotheses posed by investigators and the limitations of each exposure metric. The value of environmental measurements, such as surface and toy wipes and indoor air or house dust samples, deserves further investigation. Emphasis on personal rather than environmental sampling in conjunction with urine or blood sampling is likely to be most effective at classifying exposure. For infants and young children, ease of urine collection (possible for extended periods of time) may make these samples the best available approach to capturing exposure variability of nonpersistent pesticides; additional validation studies are needed. Saliva measurements of pesticides, if feasible, would overcome the limitations of urinary metabolite-based exposure analysis. Global positioning system technology appears promising in the delineation of children's time-location patterns.