While synthetic siRNA libraries are powerful tools for functional genomic screens, off-target effects mediated by siRNA seed interactions with the 3' UTR of unintended targets can result in false positives. Given the frequency of off-target effects in some assays, the development of hit validation/stratification strategies is imperative. In the following study we have compared two strategies for identification of high confidence hits: 1) a multiple reagent approach where two or more individual siRNAs induce the same phenotype and 2) a chemical modification approach where hit confirmation is achieved using pools of siRNA that contain specificity enhancing modifications. A comparison of these two strategies (using a collection of primary hits generated from a cell viability screen) reveals significant overlap between the high confidence hits identified. However, for low confidence hits, i.e. where a single siRNA induces a phenotype, the concern is that an important hit will be missed. To determine if the phenotype is due to gene targeting or a seed-mediated off-target effect, a chimeric approach was used whereby a gene-specific seed sequence is introduced into a non-targeting siRNA scaffold. Together, these data provide well-defined approaches for prioritization of hits derived from RNAi screens.