The Individual and National Survey on Food Consumption (INCA1), conducted in 1999 on a sample of 3,000 individuals representative of the French population, is the only nationwide source of data on the consumption of tap water for drinking. These data are presented here by distinguishing, on the one hand, the consumption of "unheated" water—which must be taken into account when assessing exposure to pathogenic microorganisms—and, on the other hand, "total" water consumption (including water for coffee and tea), which is suitable for estimating exposure to toxins. The first section describes consumption and examines its potential determinants: weight, age, sex, socio-professional category, place of residence, housing density, and consumption of alternative beverages. An appendix provides detailed values by region, age, and sex, as well as statistical distribution models. The study shows that the volumes of tap water consumed are lower than those reported in the literature. Thirty percent of the population does not drink unheated tap water. For consumers (69% of the population), the average volume consumed is 0.4 L per day. Regarding total water consumption, these figures are 6% of non-consumers and 0.5 L per day, respectively. Consumption varies considerably, however, depending on age and region. The effect of locally distributed water quality on consumption—not reported here—limits the use of these data to ordinary conditions, excluding use at too fine a geographic scale and/or in exceptional circumstances.