The PharmPrint methodology, as modified and implemented by Deanda and Stewart, was prospectively evaluated for use as a virtual high-throughput screening tool by applying it to the design of target-focused arrays. To this end, PharmPrint quantitative structure-activity relationship (QSAR) models for the prediction of AKT1, Aurora-A, and ROCK1 inhibition were constructed and used to virtually screen two large combinatorial libraries. Based on predicted activities, an Aurora-A targeted array and a ROCK1 targeted array were designed and synthesized. One control group per designed array was also synthesized to assess the enrichment levels achieved by the QSAR models. For the Aurora-A targeted array, the hit rate, against the intended target, was 42.9%, whereas that of the control group was 0%. Thus, the enrichment level achieved by the Aurora-A QSAR model was incalculable. For the ROCK1 targeted array, the hit rate against the intended target was 30.6%, whereas that of the control group was 5.10%, making the enrichment level achieved by the ROCK1 QSAR model 6-fold above control. Clearly, these results support the use of the PharmPrint methodology as a virtual screening tool for the design of kinase-targeted arrays.