The sine oculis (so) and eyes absent (eya) genes are required for retinal development as mutations in either gene leads to the complete absence of the adult retina. These proteins are known to form a biochemical complex which functions as a bipartite transcription factor to activate targets necessary for eye development. Examination of null clones of either gene shows a corresponding absence of the partner protein. Together these data support a model in which so and eya regulate each other’s expression. The no-eye phenotype that characterizes the eya2 mutant is caused by the deletion of an eye specific enhancer element. Prior studies have concluded that this enhancer is the main cis-regulatory element controlling the expression of eya during eye specification. The enhancer contains a So binding site thus eya expression could be initiated and maintained by the So-Eya complex itself. Several additional So binding sites are present within the locus and a recent ChIP-seq study indicated the presence of several So peaks. We sought to identify additional cis-regulatory elements that might be responsive to the So-Eya complex but whose activity might be masked in the eya2 mutant. To this end we expressed a SoVP16 chimeric molecule in eya2 mutants and found it capable of restoring some Eya protein and partially rescuing the no-eye phenotype implying the presence of additional So responsive enhancers. We have identified multiple new cis-regulatory elements that are responsible for robust eya expression in the retina. Surprisingly, these elements are still activated in so mutants suggesting they are not regulated by So. This led us to more thoroughly examine Eya expression in so mutants. We find that the loss of Eya protein corresponds to the temporal de-repression of non-ocular genes within the eye field. We have previously shown that this de-repression ultimately results in a cell fate switch from retinal progenitor to head epidermis. Concurrently, a wave of cell death in the eye field clears retinal progenitors which, for reasons unknown, are incapable of undergoing the cell fate transformation. These two events lead to a progressive loss of Eya protein. Our examination of so mutant clones indicates that those lacking Eya protein expression ultimately differentiate into head capsule. As expected those clones which have not undergone a cell fate switch still contain Eya protein further supporting our model that so is not directly regulating eya expression in the developing retina. Finally, our data suggests that caution should be exhibited when using RD mutants to interpret regulatory interactions among RD network members as cell fate transformations can lead to false positive interactions.