Four-day composite solid food and beverage duplicate plates and 1-L samples of drinking water were collected from a stratified random sample of 80 individuals as part of the National Human Exposure Assessment Survey in Maryland. The media were obtained from each participant in up to six equally spaced sampling cycles over a year and analyzed for copper by inductively coupled plasma mass spectrometry. Copper concentrations (microg/kg) and consumption rates (kg/d) of solid food, beverage and drinking water were used to derive average daily aggregate oral intake of copper (microg/d). The mean aggregate copper intake of 263 measurements obtained from 68 people was 923.2 +/- 685.6 microg/d (mean +/- SD). Intake through solid food accounted for the majority of aggregate daily intake of copper contributing 87% on average. According to results from mixed model analysis of variance procedures, the mean log-transformed average daily copper intake in each medium except beverage exhibited significant (P < 0.05) variability among sampling cycles. Between-person variability accounted for 50% of the total variance in aggregate copper intake. As measured by the coefficient of variation, distributions of copper intake consisting of one observation per individual were more variable than the distribution consisting of the long-term average intake for each person. These results suggest that estimates of the fraction of a population at risk from chronic copper deficiency or excess copper intake can be overestimated if based upon short-term measures of copper intake. In addition, these results indicate that longitudinal information is required for accurate assessment of aggregate oral intake of copper for an individual.