The core temperature of the rainbow trout Oncorhynchus mykiss (3·5 kg) dropped to 1·0° C during the first 6 h of chilling at 0·5° C, remained stable until 24 h, and dropped significantly to 0·7° C after 39 h. Blood plasma osmolality increased and muscle moisture content decreased gradually with increasing chilling time. After 39 h of chilling, the rainbow trout experienced 40 mosmol l‐1 higher blood plasma osmolality and 2·8% less muscle moisture content compared with initial values. In the Atlantic salmon Salmo salar (5·3 kg), core temperature dropped to 1·3° C and blood plasma osmolality increased significantly during the first 6 h of chilling at 0·5° C, but remained relatively stable throughout the rest of the experimental period. After 39 h of chilling, the salmon experienced 20 mosmol l‐1 higher blood plasma osmolality and 0·5% less muscle moisture content compared with initial values. In rainbow trout muscle moisture content was inversely related to blood plasma osmolality indicating reduced seawater adaptation with increasing hours of chilling. No such relationship was observed in the Atlantic salmon. Hence, changes in plasma osmolality and muscle moisture in the Atlantic salmon do not indicate osmoregulatory failure since the new levels, once established, were maintained throughout the chilling time.

Three experiments were conducted in order to evaluate the effects of urea and trimethylamine oxide (TMAO) supplementation in diets for seawater‐farmed rainbow trout Oncorhynchus mykiss. In all experiments, commercial extruded feed was used as basal control diet, whereas experimental diets were the basal diet supplemented with either 10 g kg–1 TMAO, 10 g kg–1 urea or 10 g kg–1 TMAO and 10 g kg–1 urea. The experiments were conducted both in research units and in a commercial fish farm. Quality was evaluated in three different commercial slaughterhouses, sorting the rainbow trout into superior, ordinary or processing grade. Significantly more TMAO was observed in muscle of rainbow trout given dietary supplementation of TMAO. However, the levels were low compared with levels in marine fish in general. Trimethylamine (TMA) was not found. No increase in muscle urea content was observed in trout fed urea in the diet. Significantly reduced fat belching, seen as reduced number of fat droplets surfacing in the net pen, was observed in groups given TMAO‐supplemented feeds. At least twice as many rainbow trout with skin vesicles were observed in the control group as in the group fed urea‐supplemented diet. In addition, skin vesicles developed into open sores in the winter only in rainbow trout fed control diet. In a large‐scale trial, quality grading of almost 200 000 market‐sized rainbow trout showed significantly more (88 vs. 83%) trout of superior grade in the group fed dietary urea compared with fish fed control diet. The main reason for downgrading from superior to ordinary and processing grade was skin lesions, supporting the observations made in the experimental units.