Nerve transection experiments combined with high-speed videography and electromyography were used to characterize a prey-type dependent hypoglossal feedback system which coordinates mouth opening and tongue protraction in the common leopard frog, Rana pipiens. When feeding on small prey, sensory feedback from the tongue through the hypoglossal nerve is necessary to trigger mouth opening. If sensory feedback is prevented from reaching the brain by transection of the hypoglossal nerve, then the mouth fails to open although the feeding behavior appears otherwise normal. However, when feeding on large prey, the mouth opens normally even after the hypoglossal nerve has been transected. Thus, peripheral feedback is not necessary to trigger mouth opening when feeding on large prey, and presumably a central coordination mechanism is used. In Rana pipiens, the evolution of a new tongue morphology and a new motor pattern for feeding on small prey has been accompanied by the evolution of a novel, peripheral mechanism for coordinating tongue protraction and mouth opening. However, the primitive motor pattern for feeding on large prey and the primitive coordinating mechanism have been retained. These results imply that the neural circuits producing the different motor patterns for large and small prey are anatomically distinct at some level in the central nervous system. If they are not anatomically distinct, then sensory feedback should be necessary to trigger mouth opening regardless of which motor pattern is being expressed. While the anatomy of these distinct pathways remains to be elucidated, these results suggest that novel neural circuits may in fact underlie different behaviors even when they seem, superficially, to be relatively similar.