The Barents Sea capelin, a schooling species, undertakes extensive migrations against the predominating current north of Norway to its spawning grounds along the Norwegian and Russian coasts. Low swimming costs (i.e. high efficiency) at cruising speeds would be advantageous for capelin, enabling allocation of more energy to reproduction. A high oxygen uptake capacity may however also be beneficial, considering the difficulty of strenuous swimming against the predominant current, and to avoid predators. To investigate the swimming energetics of the Barents Sea capelin, the relationship between rate of oxygen uptake, MO 2 (mg O 2 kg − 1 h − 1 ), and swimming speed was investigated at 5 °C. A power and an exponential function described the correlation between MO 2 and swimming speed almost equally well as determined by the correlation coefficient. The power function described the relationship as MO 2 = 101.624 + 43.5 U 2.042, where U was the swimming speed in body lengths per second (bl s − 1 ) and 101.6 an estimate of the standard metabolic rate (SMR, mg O 2 kg − 1 h − 1 ). The exponential function gave MO 2 = 85.747e 0.591 U with U being the swimming speed (bl s − 1 ) and 85.7 an estimate of the SMR (mg O 2 kg − 1 h − 1 ). The active metabolic rate (AMR, mg O 2 kg − 1 h − 1 ), the oxygen uptake measured at the critical swimming speed ( U crit), was 532 mg O 2 kg − 1 h − 1 , hence resulting in a scope for activity (SFA) of 5.2 or 6.2, depending on the model for SMR estimation. The capelin was able to sustain U crit of 3 bl s − 1 , or ∼51 cm s − 1 . The relationship between cost of transport (COT) and swimming speed was characteristically U-shaped, with minimum cost of transport (COT min) between 218 and 225 mg O 2 kg − 1 km - 1 , at an optimal swimming speed ( U opt) within the range of 1.5 to 1.7 bl s − 1 . The swimming energetics of capelin was related to its migratory behaviour and the results compared to other pelagic species. Finally, the limitations involved with working with a schooling species were discussed, in addition to how choice of model for description of costs associated with swimming generally may influence parameter estimation.