Although it is generally believed that the central projections of the retina in birds are entirely crossed, using wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) as an anterograde tracer, we have found that in normal posthatched chicks there is a small ipsilateral retinofugal projection to the diencephalon and midbrain. Most of the ipsilateral fibers appear to be directed to the lateral anterior and dorsolateral anterior nuclei of the thalamus, to the pretectal region, and to the ectomammillary nucleus and the adjoining nucleus externus. Even in the best preparations the numbers of ipsilateral fibers are so small that it is hardly surprising that they have been overlooked in previous axonal degeneration and autoradiographic experiments. A significantly larger ipsilateral retinal projection develops during the second week of incubation. The ipsilaterally directed fibers can be first seen on the fifth day of incubation and their numbers appear to increase until about embryonic day 12. At this stage the projection involves substantially more fibers than at hatching and is also more extensive in its distribution; in fact, in its general organization (but not its size) it closely parallels the normal crossed retinofugal system, contributing fibers to essentially all the primary visual relay nuclei in the diencephalon and midbrain and to much of the optic tectum, where the densest projection is to its caudomedial aspect. During the second week of incubation there is also a small number of retinal fibers, which after crossing in the optic chiasm, recross the midline in the posterior and tectal commissures (and also in the tectal roof plate), before ending in the pretectal region of the ipsilateral side. In addition, there is a markedly aberrant projection from the retina into the contralateral optic nerve. Most of the ipsilateral retinal fibers are eliminated between the twelfth and sixteenth days of incubation, and by day 17 the ipsilateral projection is reduced to its mature form. The progressive reduction in the ipsilateral projection occurs at a time when it is known (from other studies) that there is an appreciable loss of retinal ganglion cells; but whether the reduction is due to neuronal death or to the selective elimination of ipsilateral axon collaterals remains to be determined. The existence of a significant ipsilateral retinofugal component early in development, probably accounts, in part, for the distinctive and persistent ipsilateral projection that occurs if one eye is removed during the first few days of incubation.