To describe the biomarker strategy that was applied to select survodutide (BI 456906), BI 456908 and BI 456897 from 19 dual glucagon receptor (GCGR)/ glucagon-like peptide-1 receptor (GLP-1R) agonists for in-depth pharmacological profiling, which led to the qualification of survodutide as the clinical development candidate. Potencies to increase cyclic adenosine monophosphate (cAMP) were determined in Chinese hamster ovary (CHO)-K1 cells stably expressing human GCGR and GLP-1R. Agonism for endogenously expressed receptors was investigated in insulinoma cells (MIN6) for mouse GLP-1R, and in rat primary hepatocytes for the GCGR. In vivo potencies to engage the GLP-1R or GCGR were determined, measuring improvement in oral glucose tolerance (30 nmol/kg) and increase in plasma fibroblast growth factor-21 (FGF21) and liver nicotinamide N-methyltransferase (NNMT) mRNA expression (100 nmol/kg), respectively. Body weight- and glucose-lowering efficacies were investigated in diet-induced obese (DIO) mice and diabetic db/db mice, respectively. Upon acute dosing in lean mice, target engagement biomarkers for the GCGR and GLP-1R demonstrated a significant correlation (Spearman correlation coefficient with p < 0.05) to the in vitro GCGR and GLP-1R potencies for the 19 dual agonists investigated. Survodutide, BI 456908 and BI 456897 were selected for in-depth pharmacological profiling based on the significant improvement in acute oral glucose tolerance achieved (area under the curve [AUC] of 54%, 57% and 60% vs. vehicle) that was comparable to semaglutide (AUC of 45% vs. vehicle), while showing different degrees of in vivo GCGR engagement, as determined by hepatic NNMT mRNA expression (increased by 15- to 17-fold vs. vehicle) and plasma FGF21 concentrations (increased by up to sevenfold vs. vehicle). In DIO mice, survodutide (30 nmol/kg/once daily), BI 456908 (30 nmol/kg/once daily) and BI 456897 (10 nmol/kg/once daily) achieved a body weight-lowering efficacy from baseline of 25%, 27% and 26%, respectively. In db/db mice, survodutide and BI 456908 (10 and 20 nmol/kg/once daily) significantly lowered glycated haemoglobin (0.4%-0.6%); no significant effect was observed for BI 456897 (3 and 7 nmol/kg/once daily). Survodutide was selected as the clinical candidate based on its balanced dual GCGR/GLP-1R pharmacology, engaging the GCGR for robust body weight-lowering efficacy exceeding that of selective GLP-1R agonists, while achieving antidiabetic efficacy that was comparable to selective GLP-1R agonism. Survodutide is currently being investigated in Phase 3 clinical trials in people living with obesity.