The glucagon-like peptide-1 receptor (GLP1R) plays a key role in glucose homeostasis and as such is an important target for type 2 diabetes therapeutics. The endogenous ligand for the receptor, GLP-1, has a short half-life in vivo and is rapidly degraded by peptidases. GLP-1R agonists in the clinic must be stabilised against such degradation by amino acid substitutions, lipidation or fusion to immunoglobins. Despite the method of stabilisation, GLP-1R agonists in the clinic retain a similar profile for gastrointestinal adverse events. Exenatide is a stable homologue of GLP-1 that was first approved in 2005, marketed under the brand name Byetta. We sought to investigate residues critical to function by performing alanine scanning of the exenatide peptide. Alanine substitutions in positions 1, 4, 7, 9 and 22 caused a complete loss of β-arrestin1 binding and receptor internalisation. These residues are conserved in the related peptides glucagon and oxyntomodulin. Calcium signalling was also attenuated, however surprisingly, cAMP signalling was unaffected for these substituted peptides in terms of maximum effect (all were >90% of reference exenatide response). The alanine substituted peptides elicited an equivalent response to native exenatide peptide (70 ng/ml insulin) in glucose stimulated insulin secretion by the rat β-cell line INS-1 832/3, and were also equivalent at controlling glucose as measured by a lean mouse (C57Bl/6, n=7) ipGTT (Blood Glucose <200 mg/dL post 2 g/kg glucose bolus, p<0.0001 vs. vehicle by one-way ANOVA). This suggests that while critical for β-arrestin and Gq-coupled signalling, these key residues are not important for β-cell function. Substitution in position 1 with phenylalanine has been reported previously 1. In our hands, [Phe1]-Exenatide was a partial agonist in cAMP accumulation assays (maximum effect 41% of reference exenatide response), highlighting the importance of cell lines used in terms of ‘method bias’. The identification of key residues involved in β-arrestin binding, receptor internalisation and G protein signalling may aid the design of biased agonists for the treatment of diabetes, selecting beneficial outcomes and minimising adverse events. Alanine-mutated exenatide peptides will be progressed to rodent taste aversion tests to determine whether signalling bias at the GLP-1R is important in cell types other than the β-cell.