The composition of snake venoms shows a high degree of variation at all taxonomic levels, and natural selection for diet has been implicated as a potential cause. Saw-scaled vipers (Echis) provide a good model for studying this phenomenon. The venoms of arthropod feeding species of Echis are significantly more toxic to natural scorpion prey than those of species which feed predominantly upon vertebrate prey. Although testing venom activity on natural prey is important for our understanding of the evolution of venom, natural prey species are often difficult to obtain in sufficient numbers for toxinological work. In order to test the viability of using cheaper and more easily available model organisms for toxicity assessments in evolutionary research, and the extent to which toxicity of arthropod-eating Echis venoms is increased to arthropods in general or targeted to certain groups, we conducted median lethal dosage (LD50) and time to death trials using the desert locust (Schistocerca gregaria) as a model arthropod, rarely consumed by wild Echis. The venoms of arthropod specialist Echis were found to be significantly more toxic to locusts than the venom of a vertebrate feeding outgroup (Bitis arietans), and one arthropod specialist venom was found to be more toxic than those species which feed upon arthropods infrequently or not at all. The venoms of arthropod specialists were also found to cause death and incapacitation faster than the vertebrate feeding outgroup. Despite some similarity of trends, there are considerable differences between the response of natural prey (scorpions) and a model arthropod (locust) to the venoms of Echis species. This suggests that when possible, natural prey rather than convenient model organisms should be used to gain an understanding of the functional significance of variation in venom composition in snakes.