The hypothesis of the present study was that Atlantic salmon (Salmo salar) would respond to large variations in supplementation of dietary pro- and antioxidants, and marine lipid, with adjustment of the endogenously synthesised antioxidants, glutathione (GSH) and ubiquinone (UQ). An experiment with 27–3 reduced factorial design (the number of cases reduced systematically from 27 (full design) to 24 (reduced design)) was conducted, where vitamins, minerals and lipid were supplemented in the diet at high and low levels. For the vitamins and minerals the high levels were chosen to be just below anticipated toxic levels and the low levels were just above the requirement (vitamin C, 30 and 1000 mg/kg; vitamin E, 70 and 430 mg/kg; Fe, 70 and 1200 mg/kg; Cu, 8 and 110 mg/kg; Mn, 12 and 200 mg/kg). For astaxanthin, the dietary levels were 10 and 50 mg/kg and for lipid, 150 and 330 g/kg. The experiment was started with post-smolts (148 (sd 17 g)) and lasted for 5 months. The only effect on GSH was a minor increase ( < 10 %) in total concentration in the liver in response to high dietary lipid. GSH redox state was not affected. UQ responded to dietary lipid, astaxanthin and vitamin E, both with regard to total concentration and redox state. Except for an effect of Fe on plasma GSH, the trace elements and vitamin C had no effect on tissue levels and oxidation state of GSH and UQ. This shows that the endogenous redox state is quite robust with regard to variation of dietary pro- and antioxidants in Atlantic salmon.