Exposure to persistent, bioaccumulating substances, through the consumption of contaminated fish is of concern in human populations. Consumption may be particularly high for subsistence, commercial, and recreational fishers, so it is important to obtain accurate consumption estimates to assess risks in these groups. The objectives of the work reported here were: to obtain estimates of fish and shellfish consumption (meals and portion size) in an occupational cohort; to determine what percentage of the consumption was from local fish; to evaluate reliability of two methods of reporting fish and shellfish consumption; and to examine risk perceptions in relation to consumption. Subjects included 99 recreational and occupational fishers in the Chesapeake Bay area, Virginia, who were recruited for a cohort study of estuary-associated syndrome. Subjects reported average fish and shellfish consumption (all species) on a weekly, monthly or yearly basis, and were asked species-specific information, which was summed. The median number of fish meals consumed a year was significantly different depending on the method used, 52 (interquartile range, IQR:24-104) (average method) and 65 (IQR:30-117) (sum of species-specific), respectively. Shellfish estimates were 24 (IQR:12-52) (average) and 47 (IQR:31-84) (sum) meals a year. Of those who consumed fish, participants reported an average meal size of 8.9+/-3.38 oz. (median 8 oz, range 4-16) with close to 70% of fish consumed self-caught and 50% from Virginia waters. Using multiple regression, occupation, and risk perceptions were found to be significantly correlated with fish consumption levels, and consumption of fish from locations under advisory.