An experiment with 0.2-kg Atlantic salmon, Salmo salar in saltwater was conducted to determine if the ®sh could grow normally, and maintain normal nitrogen (N) and mineral balance when fed a diet with the majority of the protein (75%) derived from soy-protein concentrate (SPC). The two diets contained 50% SPC and 15% ®sh meal (FM) or 60% FM as the sources of protein. No calcium phosphate was added to the diets in order to assess the availability of P from the ingredients. A second aim was to investigate if wholebody concentrations of essential elements and growth were related in individual salmon.Growth (SGR 0.88±0.89) was similar in salmon fed the two diets, and the ®sh nearly doubled their body weights during the 84 days of feeding. Feed conversion was more ecient for the FM diet (0.81 kg intake kg )1 gain) than for the SPC diet (0.89 kg kg )1 ). The intake of N was similar, faecal loss of N was lower, while the metabolic N excretion was greater in the ®sh fed the FM than the SPC diet. This resulted in a total excretion of 35.4 g N kg )1 gain for the salmon fed the FM diet and 35.5 g N for the ®sh fed the SPC diet. Both the intake, faecal and metabolic excretion of P were higher in the ®sh fed the FM diet than the SPC diet, resulting in a total excretion of 10.5 g P kg )1 gain for the FM diet and 7.2 g P for the SPC diet. Whole-body concentrations of Ca, Mg, P and Zn were lower in the ®sh fed the SPC diet, while the Ca±P ratio was decreased, both when compared with the ®sh at the start of the experiment, and the ®sh fed the FM diet. The dierences in elemental composition were ascribed to a combination of reduced availability of elements due to phytic acid and lower concentration of elements in the SPC than in the FM. No reduction in growth of individual ®sh, which could be ascribed to reduced availability of essential elements, was seen. KEY WORDS