A multi-parameter analysis was applied in order to acquire an integrated and objective assessment of the effects of recirculating aquaculture systems in rainbow trout farms on a range of welfare indicators. We compared growth, cortisol and behavior in trout previously held in a Recirculating System (RS) developed in the PEIMA (Brittany, France) with optimized water quality, with trout held in a Flow-Through System (FTS) 100% supplied with neighboring lake water. After acclimation to experimental tanks using a flow-through system, the fish were filmed 30 minutes per day, two days per week for two weeks. At the end of this observation period, we performed an Emotional Reactivity (ER) test created by the social isolation of the trout transferred into a novel environment. Trout were individually video-taped in the novel tank for 20 minutes and stress-related behaviors were automatically analyzed. Plasma cortisol was measured in undisturbed fish and in tested fish 30 minutes after the start of the ER. Learning ability was measured by group food-anticipatory activity. Mean weight was higher in the RS group, probably resulting from warmer water temperature in that group. Basal cortisol and cortisol levels measured after the acute stress did not differ between the RS and FTS fish. Group swimming activity and between-individuals dispersion registered during the observation periods were similar in both groups. Learning performance was comparable between the two groups. The ER test showed no major behavioral differences between RS and FTS trout, except for the maximum velocity recorded during the first 5 minutes which was higher for trout previously held in a FTS. The integration of performance and stress-related parameters suggests that, when water quality is kept optimum, a recirculating aquaculture system does not induce welfare impairment for rainbow trout. This study highlights the importance of integrating multi-parameter approaches for an objective and accurate trout welfare assessment.