When unfolded proteins accumulate in the endoplasmic reticulum (ER), the unfolded protein response (UPR) increases ER protein folding capacity to restore protein folding homeostasis. Unfolded proteins activate UPR signalling across the ER membrane to the nucleus by promoting oligomerisation of IRE1, a conserved transmembrane ER stress receptor. Despite significant research, the mechanism of coupling ER stress to IRE1 oligomerisation and activation has remained contested. There are two proposed mechanisms by which IRE1 may sense accumulating unfolded proteins. In the direct binding mechanism, unfolded proteins are able to bind directly to IRE1 to drive its oligomerisation. In the chaperone inhibition mechanism, unfolded proteins compete for the repressive BiP bound to IRE1 leaving IRE1 free to oligomerise. Currently, these two mechanisms respectively lack compelling in vivo and in vitro evidence required to assess their validity. The work presented here first describes in vivo experiments that identify a role of the ER co-chaperone ERdj4 as an IRE1 repressor that promotes a complex between the luminal Hsp70 BiP and the luminal stress-sensing domain of IRE1α (IRE1LD). This is then built on by a series of in vitro experiments showing that ERdj4 catalyses formation of a repressive BiP-IRE1LD complex and that this complex can be disrupted by the presence of competing unfolded protein substrates to restore IRE1LD to its default, dimeric, and active state. The identification of ERdj4 and the in vitro reconstitution of chaperone inhibition establish BiP and its J-domain co-chaperones as key regulators of the UPR. This thesis also utilises the power of Cas9-CRISPR technology to introduce specific mutations into the endogenous IRE1α locus and to screen for derepressing IRE1α mutations. Via this methodology, two predicted unstructured regions of IRE1 are found to be important for IRE1 repression. Finally, this thesis challenges recent in vitro findings concerning the direct binding mechanism.