Introduction Recirculation Aquaculture Systems (RAS) use water recycling as the base for an efficient production. Water reutilization demands cleaning techniques to remove accumulating compounds. These include mechanical filtration, biofiltration and a disinfection method. The most common disinfection methods used in RAS are UV and Ozone, both having pros and contras in terms of costs, risks, and application requirements. An alternative to these methods is hydrogen peroxide (H2O2), which, even been classified as a green oxidant, has not been commonly used and is hardly criticized due to acquisition, storage and secure manipulation issues. As H2O2 dissociates into non-toxic environmentally benign by-products including oxygen, its application offers potential savings in terms of oxygenation in RAS. Moreover, H2O2 is used to treat ectoparasitic illness with FDA approved H2O2-35% dosages between 50-1000 mg/L for 15-60 min. Minimum concentrations to which stress responses have been observed in salmonids range between 4-170 mg/L while the limits for biofilter function are around 5 mg/L [1-4]. Materials and methods In the present study, we perform serial experiments in a RAS rearing ~65 kg European seabass (Dicentrarchus labrax, weight ranging from 500-800g) to test the best application possibilities for H2O2 into a system, with concentrations between 2.4-15.8 mg/L/h. We tested two application positions: the bypass between the protein skimmer and the sump (with higher organic burden), and the flow to the rearing tanks (low organic load). We evaluated the impact on physicochemical parameters with focus on oxygen and nitrogen species concentration and the effect of the application on bacterial community of the different compartments (assessed via total bacterial counts in CFU/ml). Dosage was applied for 4 h on a daily basis and rates of decay of H2O2 on the system was assessed with Quantofix Peroxide test sticks (Sigma Aldrich, from 0-25 mg/L and 0-100 mg/L) until total removal. Ozone was switched off during dosage. Smart Digital S-DDA (Grundfos, Germany) dosage pumps were used for a controlled H2O2 application to the system (dosing capacity between 2.4 ml/h and 7.5 L/h). Stress levels of the reared fish was assessed using cortisol and glucose as blood markers. Results With the concentrations used, we accomplish a partial disinfection of the system without reaching levels affecting the bacterial flora of the biofiltration units or the health of the reared fish. We found a significant reduction of the bacterial load with each of the concentration tested. The best application position was the inlet to the protein skimmer, due to the lowest impact to the bacterial flora of the filter and the reared fish. An economically relevant increase in oxygen level of the system as well as the highest disinfection levels were achieved with 15.8 mg/L/h. In this assay, an artificial oxygen depletion was simulated before applying H2O2 to the system, and this was recovered within the first 40 min, even during a feeding period, which normally reduces the oxygen level in the water column. After application, the rates of decay varied from 30 min to 1 h, time after which, no H2O2 was detected in the system. No significant negative effects were found on fish stress markers even when cortisol (Ref: 59.8±49.5 vs. Treat:62.2±43.5) and glucose (Ref: 131.0±34.9 vs. Treat: 155.1±45.3) levels tended to increase during the experiments using increasing H2O2 doses. These markers showed a high variability. During application, the number of fish showing conspicuous low cortisol levels increased. Discussion and conclusion Organic burden in the system, fish stocking density, feeding frequency and biofilter performance were all factors influencing the required H2O2 dosage and its rate of decay in accordance to what has been published [2, 5, 6]. In the present study, we found a relatively rapid decay of H2O2 remnants, supported by the organic load typically present in a RAS rearing at commercial or semi-commercial scales. We recommend to apply H2O2 in positions of the system with high organic burden to avoid affecting biofilter function or promoting nitrite accumulation as previously published [5]. Despite previous results showing negative stress responses of D. labrax to H2O2 application [7], we did not found critical changes on cortisol or glucose related to the treatment. These authors used smaller fish (weight 120-200 g) and doses of 50 ppm/h, which are higher than in the present study. The stress in D. labrax is genetically driven and highly dependent on size/age with individuals having consistent high (439.2 ± 31.1 µg/dl ) or low (247 ± 85.1 µg/dl) cortisol concentrations and, as well as glucose, there is a circadian pattern for these parameters within this species [8, 9]. Glucose levels between 100-150 mg/dl have been reported in the literature for D. labrax [10]. With appropriate knowledge of system performance and microbiological background, H2O2 can be a positive complement to the disinfection spectrum of a RAS. It can be used alternating with common methods to avoid artificial selection of specific microbial groups and can have a positive impact in cases in which parasitic illness are present. Thus, as a disinfection method with a plus for its oxygenation influence, H2O2 can enrich the disinfection portfolio of aquaculture facilities.