Since the discovery of RNA interference almost 15 years ago, researchers have tried to devise effective ways to get these short gene-silencing transcripts inside cells. But techniques for in vivo delivery have fallen short. For therapies that require long-term gene knockdown, DNA viruses that insert the microRNA (miRNA) genes into the target cell’s genome were the best bet. But that approach produced large numbers of miRNAs, which overloaded the miRNA-processing proteins in the nucleus and resulted in cellular toxicity. For short-term therapies, mature miRNAs encapsulated in liposomes or conjugated to cholesterol were used, but they have been difficult to target to a specific cell type.