Abstract Background/Introduction Standard-of-care anticoagulants are associated with increased bleeding risk. Genetic/pharmacological data suggest coagulation factor XI (FXI) inhibition can prevent thrombosis with minimal increased bleeding risk. We present preclinical and first-in-human (FIH) data on REGN9933, a FXI-binding monoclonal Ab. Purpose To characterise REGN9933 epitope binding sites, affinities for FXI and activated FXI (FXIa), and effect on high molecular weight kininogen (HMWK)’s interaction with FXI. Also, to assess safety, tolerability and pharmacodynamics in healthy human participants (pts). Methods Epitope binding sites mapped with cryogenic electron microscopy, binding affinities determined with plasmon resonance techniques, effect on thrombin generation assessed with a thrombin generation assay, and impact on HMWK:FXI interaction analysed using ELISAs. In the FIH, Ph 1, randomised, double-blind, single-centre study, pts received a single dose of REGN9933 intravenously (IV; 3–300 mg) or subcutaneously (SC; 100 mg & 300 mg), or placebo (PBO). Primary endpoint: incidence and severity of treatment-emergent AEs (TEAEs). Other endpoints: effect of REGN9933 on activated partial thromboplastin time (aPTT) and prothrombin time (PT) to assess intrinsic and extrinsic pathway-triggered coagulation, respectively; and FXI activity. Results REGN9933 was found to bind the FXI apple 2 domain, with subnanomolar binding affinities for FXI and FXIa at 37°C, resulting in reduced thrombin generation from the intrinsic (not extrinsic) pathway. REGN9933 competed with HMWK for FXI binding in a dose-related manner. In the Ph 1 study, 56 pts received REGN9933 (IV, n=30; SC, n=12) or PBO (n=14); 42.9% and 21.4% had ≥1 TEAE, respectively (most common with REGN9933: headache, 12%; oropharyngeal pain, 5%). No serious/severe TEAEs, or TEAEs of special interest (inc. moderate/severe bleeding) reported. Laboratory tests revealed no clinically meaningful differences in the number/type of treatment-emergent potentially clinically significant values with REGN9933 vs PBO. REGN9933 resulted in dose-dependent prolongation of aPTT; aPTT mean fold change from baseline was highest with IV 300 mg after 8 hrs (2.7) and remained >2 for 36 days (Fig 1). There was no detectable effect on PT or clinically meaningful effect on bleeding time. REGN9933 supressed FXI activity in a dose-dependent manner, with suppression to approx. the lower limit of assay quantification (5%) with IV ≥30 mg (Fig 2). Conclusion(s) REGN9933 binds the FXI apple 2 domain and inhibits HMWK:FXI interaction, preventing intrinsic pathway-triggered FXI activation. FIH data showed dose-dependent inhibition of the intrinsic coagulation pathway by REGN9933, without affecting the extrinsic or common pathways. Pharmacodynamic and safety findings support continued development of REGN9933 as an anticoagulant that may carry less bleeding risk than currently approved agents.