We have investigated with high resolution the timing of retinal precursor cell commitment to specific differentiated fates, using an in ovo modification of the in vitro "window-labeling" technique (A. M. Repka and R. Adler, J. Histochem. Cytochem. 40, 947-953, 1992a). The method involves an initial injection of tritiated thymidine into chick embryos, followed a specific number of hours later by an injection of bromodeoxyuridine (BrDU); cells born during this period are identified by being labeled with thymidine but not with BrDU. We used this method to determine, in a narrow region adjacent to the choroid fissure, the fate of cells born during defined 5-hr intervals between Embryonic Days (ED) 4-8. All the cohorts gave rise to heterogenous differentiated populations, indicating that time of cell birth is not a major cell fate determinant. A progressive restriction in the developmental potential of precursor cells, however, was suggested by the observed decrease in the number of different populations generated during each 5-hr period from ED 4 to 8, and supported also by dissociated cell culture experiments investigating the fate of cells born at different developmental stages. Microenvironmental influences were tested in vitro using cells windowed-labeled in ovo for 5 hr on ED 5. After spending at least 72 hr within the retina before their isolation for culture, these cells mimicked their in vivo fate, giving rise predominantly to nonphotoreceptor neurons; a completely different behavior was observed when the cells were isolated after shorter exposures to the retinal microenvironment, when they gave rise predominantly to photoreceptors. Together with data demonstrating that differential cell death cannot account for these results, our results are consistent with the hypothesis that cell fate determination occurs after the time of terminal mitosis.