Juvenile (0+ year) perch Perca fluviatilis, held in tanks (1 m3) for 8 months under constant temperature (17° C) and light (12L : 12D) conditions, were weighed monthly. Individually marked fish from the tanks were placed in aquaria in groups of four and tested for feeding success, assessed as individual share of a group meal in three identical short‐term aquarium experiments separated in time (January, May and June). The groups consisted of the same individuals in each of the three experiments. In tanks, growth (thermal growth coefficient, GT) was used as a measurement of feeding success. Individual GT in tanks was consistent over time, and individual share of a group meal in aquaria was consistent between consecutive months. There was no connection between individual feeding success in aquaria and GT in tanks in May and June, and in January the factors were negatively correlated. The condition factor at the start of aquarium experiments and feeding success in aquaria were negatively correlated in June and tended to be so also in May. Boldness in aquaria was negatively correlated with growth in tanks. Thus, individuals that were successful in one environment seemed to be less successful in the other. There was a seasonal change in GT, with an increase from March to April and from April to June despite constant day length and temperature.