Stem cells have the unique ability to switch from a proliferative cell state in which they reproduce themselves to a differentiated cell state that allows them to carry out the functions of fully mature cells. The stem cell field has made considerable effort to address the underlying mechanisms behind this switch. Most effort has been focused on a class of proteins that determine whether our genetic material—DNA—generates a copy of itself in the form of messenger RNA (mRNA), a process called transcription. Some factors that regulate transcription activate the production of mRNAs encoding proteins that promote the proliferative and undifferentiated cell state. Others increase the production of messenger mRNAs that encode inhibitors of differentiation and cell proliferation. While this transcription factor-centric view of stem cells explains some aspects of stem cells, it is not sufficient to explain many of their behaviors. We hypothesize that another important mechanism controlling the stem vs. differentiation switch is regulation of mRNA decay. By increasing the rate of decay of specific mRNAs, their steady-state level will be reduced, leading to lower levels of their encoded proteins being made. Likewise, decreasing mRNA decay will lead to increased mRNA level and increased protein production. We hypothesize that a particular RNA decay pathway called nonsense-mediated RNA decay (NMD) is critical for stem cell decisions. NMD recognizes specific sequence features in RNAs and then activates enzymes to degrade them. Our project revolves around two main ideas. First, that NMD promotes the stem-like state by preferentially degrading messenger RNAs that swing the balance towards the stem-like state. Second, that NMD must be downregulated in magnitude to allow stem cells to differentiate. During the latest progress period, we obtained substantial evidence for both of these hypotheses. With regard to the first hypothesis, we used genome-wide approaches to identify hundreds of mRNAs that are regulated by NMD in human embryonic stem (ES) cells. We found that one of the main classes of mRNAs regulated by NMD is those encoding signaling proteins. We examined one particular signaling pathway—called the TGFβ pathway—and found that NMD is a potent inhibitor of this pathway in human ES cells. This was of interest given that TGFβ pathway is critical for differentiation of human ES cells into endoderm, one of the three initial tissue types in mammalian embryos. We found that the magnitude of NMD is dramatically inhibited when human ES cells differentiated into endoderm, which raised the possibility that this downregulatory event allows for increased TGFβ signaling and consequent endoderm differentiation. During the progress period, we obtained several lines of evidence in support of this model. Indeed, we found that NMD downregulation is both necessary and sufficient for human ES cells to differentiate into endoderm. As further support, we found that induced pluripotent cells (iPSCs) from patients with mutations in a key NMD gene, UPF3B, have a tendency to spontaneously differentiate. UPF3B is a particularly interesting gene given that loss of UPF3B causes intellectual disability in both mice and humans (the former from our unpublished research). Patients with mutations in UPF3B also often suffer from schizophrenia, autism, and/or attention-deficit/hyperactivity disorder. Thus, the study of NMD may provide insight into a wide spectrum of cognitive and psychological disorders.