Emerging data suggests that some types of nanoparticles (NPs) are toxic to fish, and given the well-known toxicity of dissolved metals, there are also concerns about whether metal-containing NPs present a similar or different hazard to metal salts. In this study, juvenile rainbow trout were exposed in triplicate to either a control, 20 or 100 μg l⁻¹ of either Cu as CuSO₄ or Cu-NPs (mean primary particle size, 87±27 nm) in a semi-static aqueous exposure regime. Fish were sampled at days 0, 4, and 10 for tissue trace elements, haematology, and biochemistry. By day 4, fish from the 100 μg l⁻¹ Cu as CuSO₄ treatment showed 85% mortality (treatment subsequently terminated) compared to 14% in the 100 μg l⁻¹ Cu-NP exposed fish. Mortality at day 10 was 4, 17, 10, and 19% in the control, 20 μg l⁻¹ Cu as CuSO₄, 20 and 100 μg l⁻¹ Cu-NP treatments, respectively. Copper accumulation was seen in the gills of fish from all Cu treatments, and was statistically significant in both CuSO₄ treatments at day 4 and all Cu treatments at day 10 compared to controls. No statistically significant Cu accumulation was seen in the spleen, brain or muscle of fish from any treatment, although an elevation in intestinal Cu was seen in the high Cu-NP treatment throughout. There were some transient changes in haematology and depletion of plasma Na⁺ that was treatment-related, with some differences between the nano form and metal salt, but Cu-NPs were not overtly haemolytic. A 6-fold decrease in branchial Na⁺/K⁺-ATPase activity in all Cu treatments (compared to controls), depletion of plasma and carcass ion concentrations suggest that Cu-NPs are an ionoregulatory toxicant to rainbow trout. Statistically significant decreases in Na⁺/K⁺-ATPase activity were also seen in the brains and intestine, and whilst there was no material-type effect in the former, this was only observed in the gut of fish exposed to 100 μg l⁻¹ Cu-NPs. There were material-dependent changes in tissue thiobarbituric acid reactive substances (TBARS), and in the gill the Cu-NPs caused a larger (though non-significant compared to control) increase in TBARS than the equivalent metal salt treatment (the latter actually being significantly reduced compared to all other treatments). Overall, these data show that Cu-NPs have similar types of toxic effects to CuSO₄, which can occur at lower tissue Cu concentrations than expected for the dissolved metal.