In general, copper is primarily an osmoregulatory toxicant to fish and Cu toxicity is thought to be related to the rate of sodium loss. Looking at a previous research it is striking that gibel carp, Carassius auratus gibelio, do not seem so susceptible to the first ionoregulatory shock phase of Cu exposure, but rather build up physiological disturbances slowly until mortality occurs. Since it was noted that gibel carp experience severe hypoxia under Cu exposure, we hypothesised that, besides the Na loss, the slow depletion of liver glycogen stores contributed equally to the collapse of physiological integrity. It is clear from our results that glycogen stores are being depleted in Cu exposed fish and that dead fish suffered more extensive glycogen losses compared to surviving fish, with liver glycogen levels of 125+/-8 mg g(-1) in dead fish compared to 230+/-13 mg g(-1) in surviving fish. However, changes in liver glycogen did not contribute significantly to mortality, while changes in whole body sodium and the rate of sodium loss did. Whole body Na levels dropped from 1111+/-48 microg g(-1) dry weight in control fish to 850+/-54 microg g(-1) in surviving fish and to 607+/-24 microg g(-1) in fish that had died resulting in Na loss rates of 1.25+/-0.22 microg g(-1) h(-1) and 3.39+/-0.19 microg g(-1) h(-1) in surviving and dead fish respectively. Our results support the finding that the rate of Na loss largely determines Cu toxicity in fish, even in resistant species.