higher rates of fish consumption (Rupp et al. 1980). In addition, specific ethnic and cultural subgroups, as well as recreational fishermen, can have increased exposures (EPA 1997a). Population-based estimates of MeHg exposure in the United States have been made on the basis of dietary assessment studies, which provide information on fish consumption by species and by portion size. The combination of intake frequency by species and portion size by species for each individual consumer provides an estimate of the average mass of fish consumed (in grams per day). Summaries of such studies giving national data are provided in EPA's report to Congress (EPA 1997a). Another such dietary assessment study was conducted in New Jersey (Stern et al. 1996). To estimate population-based MeHg exposure from such studies, the gram-per-day amount of each species consumed by each individual is multiplied by the characteristic MeHg concentration of each species (microgram per gram) and then is summed across species to give the average intake of MeHg by each individual (microgram/day). The distribution of individual intakes for the study sample can then provide an estimate of MeHg intake in the underlying population. Uncertainties in such assessments include those in recall and recording of intake frequency and portion size, misidentification of the species consumed, extrapolation of short-term dietary studies to long-term average exposure, and the outdated and incomplete national database on average MeHg concentrations of different fish species. Estimates also typically vary depending on the length of time over which the fish-intake data was obtained (e.g., 1-day recall versus 1-week recall). These uncertainties are discussed by EPA (1997a) and Stern et al. (1996). Table 2-3 presents the EPA (1997c) analysis of MeHg intake for the general population and for the population of women of childbearing age based on fish-consumption data for month-long consumption. Estimates based on intake from such data are generally lower than those based on 1-day dietary data. Table 2-3 also presents data from New Jersey based on a 7-day recall survey. These data, along with the study by Rupp et al. 1980, suggest that the population in that region of the United States has higher intakes than the U.S. population in general. Estimates of population exposure and risk based on the average exposure of the U.S. population might, therefore, underestimate exposure to large subpopulations. Upon completion, data from Continuing Survey of Food Intakes by Individuals (CFSII) and National Health and Nutrition Examination