The addition of a third eye primordium to the forebrain region of a Rana pipiens embryo invariably results in the development of a series of periodic, mutually exclusive eye-specific bands in tectal lobes dually innervated by the host and supernumerary fibers. A number of investigators have proposed that such source-specific segregation arises as a compromise between two mechanisms that are normally involved in retinotectal map formation: one which is dependent on cell surface affinities to align the map and produce a rough retinotopy and a second that “fine tunes” the map by stabilizing adjacent terminals from neighboring retinal ganglion cell bodies at the expense of terminals from non-neighboring cells. In this study we have tested the idea that this second “fine-tuning” mechanism is dependent on neural activity by blocking impulse activity in the optic nerves of three-eyed tadpoles. To assess the requirement for activity on the formation of bands, both normal optic nerves of 17 three-eyed tadpoles were crushed intraorbitally. Two weeks after this operation, the supernumerary retinal projection had debanded and spread to cover the entire tectum in a continuous fashion. By 4 weeks, however, the host optic fibers regenerated back to the tecta and began to form segregated stripes with the fibers from the third eye. Six to 7 weeks after the optic nerve crush the periodic pattern of eye-specific segregation characteristic of dually innervated tecta was again pronounced. When activity in all three optic nerves was eliminated with tetrodotoxin (TTX; embedded in a slow release plastic) during the last 3 weeks of this process, the fibers from the two competing eyes failed to segregate and, instead, formed two completely overlapping, continuous projections across the tectal surface. To test for the requirement of activity in the maintenance of segregation, we also subjected three-eyed tadpoles without optic nerve crush to TTX blockade for 2, 3, and 4 weeks. Animals sacrificed at 2 weeks show overlap of the projections in the rostral tectum but distinct interdigitating stripes in other regions of these lobes. After 3 weeks of blockade, segregation of the projections was less distinct in the central tectum as well. After 4 weeks of TTX blockade the terminals from both eyes spread to form continuous overlapping projections throughout the tectum. Examination of well isolated, individual retinal ganglion cell terminal arbors during this period reveals that they occupy a significantly greater area of tectum following the TTX treatment.