To date, most researchers have used dietborne metal concentrations rather than daily doses to define metal exposure and this has resulted in contradictory data within and between fish species. It has also resulted in the impression that high concentrations of dietborne Cu and Zn (e.g.>900 mg kg⁻¹ dry diet) are relatively non-toxic to fish. We re-analyzed existing data using rations and dietborne metal concentrations and used daily dose, species and life stage to define the toxicity of dietborne Cu and Zn to fish. Partly because of insufficient information we were unable to find consistent relationships between metal toxicity in laboratory-prepared diets and any other factor including, supplemented metal compound (e.g. CuSO₄ or CuCl₂), duration of metal exposure, diet type (i.e. practical, purified or live diets), or water quality (flow rates, temperature, hardness, pH, alkalinity). For laboratory-prepared diets, dietborne Cu toxicity occurred at daily doses of >1 mg kg⁻¹ body weight d⁻¹ for channel catfish (Ictalurus punctatus), 1–15 mg kg⁻¹ body weight d⁻¹ (depending on life stage) for Atlantic salmon (Salmo salar) and 35–45 mg kg⁻¹ body weight d⁻¹ for rainbow trout (Oncorhynchus mykiss). We found that dietborne Zn toxicity has not yet been demonstrated in rainbow trout or turbot (Scophthalmus maximus) probably because these species have been exposed to relatively low doses of metal (<90 mg kg⁻¹ body weight d⁻¹) and effects on growth and reproduction have not been analyzed. However, daily doses of 9–12 mg Zn kg⁻¹ body weight d⁻¹ in laboratory-prepared diets were toxic to three other species, carp Cyprinus carpio, Nile tilapia Oreochromis niloticus, and guppy Poecilia reticulata. Limited research indicates that biological incorporation of Cu or Zn into a natural diet can either increase or decrease metal bioavailability, and the relationship between bioavailability and toxicity remains unclear. We have resolved the contradictory data surrounding the effect of organic chelation on metal bioavailability. Increased bioavailability of dietborne Cu and Zn is detectable when the metal is both organically chelated and provided in very low daily doses. We have summarized the information available on the effect of phosphates, phytate and calcium on dietborne Zn bioavailability. We also explored a rationale to understand the relative importance of exposure to waterborne or dietborne Cu and Zn with a view to finding an approach useful to regulatory agencies. Contrary to popular belief, the relative efficiency of Cu uptake from water and diet is very similar when daily doses are compared rather than Cu concentrations in each media. The ratio of dietborne dose:waterborne dose is a good discriminator of the relative importance of exposure to dietborne or waterborne Zn. We discuss gaps in existing data, suggest improvements for experimental design, and indicate directions for future research.