Given the remarkable recent progress in gene therapy-based treatments for retinal disease, there is an urgent need for the development of new approaches to quantitative design and analysis of photoreceptor-specific promoters. In this study, we determined the relative binding affinity of all single-nucleotide variants of the consensus binding site of the mammalian photoreceptor transcription factor, Crx. We then showed that it is possible to use these data to accurately predict the relative binding affinity of Crx for all possible 8 bp sequences. By rationally adjusting the binding affinity of three Crx sites, we were able to fine-tune the expression of the rod-specific Rhodopsin promoter over a 225-fold range in living retinas. In addition, we showed that it is possible to fine-tune the activity of the rod-specific Gnat1 promoter over ∼275-fold range by modulating the affinity of a single Crx-binding site. We found that the action of individual binding sites depends on the precise promoter context of the site and that increasing binding affinity does not always equate with increased promoter output. Despite these caveats, this tuning approach permits quantitative engineering of photoreceptor-specific cis-regulatory elements, which can be used as drivers in gene therapy vectors for the treatment of blindness.