Tool use was once considered the sole domain of humans. Over the last 40 years, however, it has become apparent that tool use may be widespread across the animal kingdom. Pioneering studies in primates have shaped the way we think about tool use in animals, but have also lead to a bias both in terms of our expectations about which animals should be capable of using tools and the working definition of tool use. Here I briefly examine tool use in terrestrial animals and consider the constraints of the current working definition of tool use in fishes. Fishes lack grasping limbs and operate underwater where there are clear constraints with respect to the physics of tool use that differ dramatically from the terrestrial environment. I then examine all of the documented accounts of tool use in fishes. The review reveals that tool use seems to be confined to a limited number of fish taxa, particularly the wrasse, which may show similarities with the greater than expected number of examples of tool use in primates and corvids amongst mammals and birds, respectively. As fish are seldom studied as intensely as birds and mammals, there is a clear need for further observation of tool use in fishes. It is likely that further examples will be unveiled allowing us to perform comparative analyses of the evolution of tool use in fish.