It would often be practical to starve the cultivated fishes over the weekends, e.g. to save in labour costs. We evaluated the possibility that domesticated juvenile rainbow trout (Oncorhynchus mykiss) could compensate for the lost growth of the 2-day weekend starvation by either hyperphagic response and/or by lower feed conversion ratio, compared to the fish fed every day in an 8-week experiment. Rainbow trout (initial weight c. 30 g, temperature 16 degrees C) starved during the weekends were able to increase feed intake during the weekdays clearly above the intake of the control fish after the first two weekends, also seen as an increase of the compensation coefficients over the last four weeks of the experiment. However, the control fish had significantly greater total absolute feed intake than the fish in the treatment group inducing significantly higher specific growth rate and final weight (176.7 +/- 4.7 g) compared to the treatment group (153.1 +/- 8.4 g) at the end of the experiment. The coefficient of variation of weight did not differ between the two groups, but the feed conversion ratio was significantly higher in the 2-day starving group (0.93 +/- 0.03) than in the control group (0.88 +/- 0.01). Growth compensation was only partial, and our data indicated clear differences in the capacity for compensation between the tanks. The tank-wise compensation capacity appears to be rather persistent, seen as a significant positive correlation in the compensation coefficients between the first and last four weeks of the experiment. Body moisture and hepatosomatic index were significantly higher in the 2-day starving groups than in the control group, but there were no treatment differences in stomach capacity (weight or volume), liver weight or in the relative amount of visceral fat. The lack of the increase in stomach capacity in the treatment group fish was hypothesized to be a consequence of selective breeding for several decades and being an unnecessary trait in farmed fishes supplied feed constantly. We conducted also a 3-week follow-up experiment to test the effects of feeding frequency (continuous feeding vs. feeding twice per day) on growth responses in rainbow trout experiencing the weekend starvation, and these results confirmed that the lack of full growth compensation in the first experiment was not due to feeding frequency. Our results suggest that weekend starvation cannot be recommended for domesticated rainbow trout without negatively affecting their growth.