Hydrogen isotope (δ2H) measurements of consumer tissues in aquatic food webs are useful tracers of diet and provenance and may be combined with δ13C and δ15N analyses to evaluate complex trophic relationships in aquatic systems. However, δ2H measurements of organic tissues are complicated by analytical issues (e.g., H exchangeability, lack of matrix‐equivalent calibration standards, and lipid effects) and physiological mechanisms, such as H isotopic exchange with ambient water during protein synthesis and the influence of metabolic water. In this study, δ2H (and δ15N) values were obtained from fish muscle samples from Lake Winnipeg, Canada, 2007–2010, and were assessed for the effects of species, feeding habits, and ambient water δ2H values. After lipid removal, we used comparative equilibration to calibrate muscle δ2H values to nonexchangeable δ2H equivalents and controlled for H isotopic exchange between sample and laboratory ambient water vapor. We then examined the data for evidence of trophic δ2H enrichment by comparing δ2H values with δ15N values. Our results showed a significant logarithmic correlation between fork length and δ2H values, and no strong relationships between δ15N and δ2H. This suggests the so‐called apparent trophic compounding effect and the influence of metabolic water into tissue H were the potential mechanisms for δ2H enrichment. We evaluated the importance of water in controlling δ2H values of fish tissues and, consequently, the potential of H isotopes as a tracer of provenance by taking account of confounding variables such as body size and trophic effects. The δ2H values of fish appear to be a good tracer for tracking provenance, and we present a protocol for the use of H isotopes in aquatic ecosystems, which should be applicable to a broad range of marine and freshwater fish species. We advise assessing size effects or working with fish of relatively similar mass when inferring fish movements using δ2H measurements.