Aquatic food webs are subsidized by allochthonous resources but the utilization of these resources by consumers can be difficult to quantify. Stable isotope ratios of hydrogen (deuterium:hydrogen; deltaD) potentially distinguish allochthonous inputs because deltaD differs between terrestrial and aquatic primary producers. However, application of this tracer is limited by uncertainties regarding the trophic fractionation of deltaD and the contributions of H from environmental water (often called "dietary water") to consumer tissue H. We addressed these uncertainties using laboratory experiments, field observations, modeling, and a literature synthesis. Laboratory experiments that manipulated the deltaD of water and food for insects, cladoceran zooplankton, and fishes provided strong evidence that trophic fractionation of deltaD was negligible. The proportion of tissue H derived from environmental water was substantial yet variable among studies; estimates of this proportion, inclusive of lab, field, and literature data, ranged from 0 to 0.39 (mean 0.17 +/- 0.12 SD). There is a clear need for additional studies of environmental water. Accounting for environmental water in mixing models changes estimates of resource use, although simulations suggest that uncertainty about the environmental water contribution does not substantially increase the uncertainty in estimates of resource use. As long as this uncertainty is accounted for, deltaD may be a powerful tool for estimating resource use in food webs.