Abstract There are probably many reasons for the widespread belief that temnospondyls and other early stegocephalians were largely restricted to freshwater, but three of the contributing factors will be discussed below. First, temnospondyls have been called amphibians (and thought to be more closely related to extant amphibians than to amniotes). Some authors may have simply concluded that, like extant amphibians, temnospondyls could not live in oceans and seas. Second, under some phylogenies, temnospondyls are more closely related to anurans (and possibly urodeles) than to gymnophionans and could be expected, for parsimony reasons, to share the intolerance of all extant amphibians to saltwater. Similarly, ‘lepospondyls’ are often thought to be more closely related to gymnophionans than to anurans, and could also be expected to lack saltwater tolerance. Third, extant lungfishes live exclusively in freshwater, and early sarcopterygians have long been thought to share this habitat. These interpretations probably explain the widespread belief that early amphibians and early stem-tetrapods were largely restricted to freshwater. However, these three interpretations have been refuted or questioned by recent investigations. A review of the evidence suggests that several (perhaps most) early stegocephalians tolerated saltwater, even although they also lived in freshwater.