Speed of locomotion plays an important role in an animal’s biology and ecology and is of particular interest in aquatic animals. Determining cruising speeds of shark—for which such data are scarce—may help to better understand their interactions with prey, the size of the home ranges they maintain, their energetic costs and how they interact with their environment through sensory perception. In this study, the cruising speeds of a range of different shark species were measured using stereo-baited remote underwater video systems (stereo-BRUVS). The relationship between cruising speeds and fork length, species, order, habitat, trophic level, temperature and tail shape was then modelled. Fork length and species best explained the cruising speeds of the eight species of shark studied: Carcharhinus amblyrhynchos, Carcharhinus albimarginatus, Carcharhinus obscurus, Furgaleus macki, Carcharhinus obesus, Mustelus antarcticus, Heterodontus portusjacksoni and Parascyllium variolatum. This linear model had a slope that did not differ statistically from that of the theoretical model proposed by Weihs (Scale effects in animal locomotion, Academic Press, New York, pp 333–338, 1977), which suggests that the relationship between cruising speed and length appears to be dominated by energetics. The results suggest that existing allometric estimates of cruising speeds can be improved by defining cruising speeds for each species as a function of length. Currently, literature presents cruising speed data for only a few species of shark; therefore, we provide a second, generalised model, which predicts cruising speed as a function of length and tail shape. The length + tail shape model was selected based on its generality and accuracy in estimating shark cruising speeds obtained from acoustic tags. This length + tail shape model was significantly better than a length only model; it explained a further 76 % of the variation in cruising speed derived from stereo-BRUVS and acoustic tagging data than a length only model. The more accurate prediction of the length + tail shape model is most likely because tail shape is correlated with a number of ecological factors.