Knowing the target sequence of a DNA-binding protein is vital in obtaining fundamental characteristics of the protein and evaluating properties of the protein-DNA interaction. For example, novel homing endonucleases cannot be proven to be functional until a predicted target site is tested. Unfortunately, target site prediction is not always easy, or even possible, depending on the amount of sequence data available. Here we describe a modification of SELEX using yeast surface display that can quickly and inexpensively resolve DNA-binding targets in high throughput for proteins without any prior assumptions or knowledge regarding the target site. This protocol is easily integrated into the yeast surface display pipeline and is leveraged by the expansive number of existing tools for both SELEX and yeast surface display.