We present a straightforward process for the discovery of novel back pocket-binding fragment molecules against protein tyrosine kinases. The approach begins by screening against the nonphosphorylated target kinase with subsequent counterscreening of hits against the phosphorylated enzyme. Back pocket-binding fragments are inactive against the phosphorylated kinase. Fragment molecules are of insufficient size to span both regions of the ATP binding pocket; thus, the outcome is binary (back pocket-binding or hinge-binding). Next, fragments with the appropriate binding profile are assayed in combination with a known hinge-binding fragment and subsequently with a known back pocket-binding fragment. Confirmation of back pocket-binding by Yonetani-Theorell plot analysis progresses candidate fragments to crystallization trials. The method is exemplified by a fragment screening campaign against vascular endothelial growth factor receptor 2, and a novel back pocket-binding fragment is presented.