It has been well established that the shRNA library has a significant advantage for screening the important genes involved in the interested biological pathways. Currently, the available libraries mainly target the known protein genes in human and mouse. With the expanding roles of IncRNA in biology, there is a great demand to design shRNAs targeting these non-coding RNAs. In this regard, a completely random shRNA library targeting all the genes with known or unknown sequences is of priority. Here we provide a practical workflow for construction of such a random shRNA library. In the novel shRNA library, there are about tens of different shRNAs targeting one gene, and thus significantly avoids the off-target effects. (C) 2012 Elsevier Inc. All rights reserved.