The accepted experimental method of gene silencing by RNA interference (RNAi) in mammals is the introduction of small interfering RNAs (siRNAs) into cells of interest. The canonical siRNA comprises a 21 bp, double-stranded RNA molecule with the strands offset to create 2 nt overhangs at the 3′ ends. However, these siRNAs suffer from various disadvantages, including RNase susceptibility, immunogenicity to the host, and nonspecific selection of the active strand, resulting in decreased potency and potential off-target effects. Researchers have been working on improving siRNA design by altering the structure or introducing chemical modifications. If the promise of RNAi-based therapeutics is to be fulfilled, more potent and target-specific siRNAs need to be developed, and research in this field is prolific. This article investigates siRNA processing and RISC assembly, comparing different canonical siRNAs to their sequence-matched Dicer-substrate siRNAs (DsiRNAs). In their comparison, the authors define the molecular basis of DsiRNA contribution to improved performance in various stages of the RNAi pathway.