Water quality criteria for metals have largely been based on toxicity tests in which organisms are exposed to dissolved metals, often at concentrations that exceed by several orders of magnitude those which are found even in contaminated waters; these tests assume that ecological impacts can be extrapolated from acute toxicity assessments. We have shown that when marine copepods are exposed to metals (Ag, Cd, Hg, Se, Zn) through their diet the reproductive capacity decreases by up to 75% because fewer eggs are produced and the hatching success of the eggs which are produced is diminished. These sublethal effects occur at metal concentrations that are 2-3 orders of magnitude below acutely toxic concentrations. Metals accumulated in copepods by trophic transfer deposit in internal tissues, and only modest increases above background levels can depress reproductive capability. Metals accumulated from the dissolved phase deposit primarily on external surfaces, where they have negligible effects at environmentally realistic concentrations. Toxicity tests must consider the bioaccumulation of metals from dietary exposure in order to account for sublethal effects at metal concentrations that are comparable to those in contaminated waters.