To compare the effects on von Willebrand factor release of the mixed vasopressin type 1a and type 2 receptor agonist arginine vasopressin and the selective vasopressin type 1a receptor agonist FE 202158, [Phe2,Ile3,Hgn4,Orn(iPr)8]vasopressin, at doses required for the treatment of septic shock. Prospective, randomized, controlled laboratory experiment. University animal research facility. Twenty-four chronically instrumented sheep. After a 5-day recovery from instrumentation, sheep were randomly assigned to receive a single intravenous bolus of the selective vasopressin type 2 receptor agonist desmopressin (1 nmol·kg(-1)) or continuous intravenous infusions of arginine vasopressin (3 pmol·kg(-1)·min(-1)), the selective vasopressin type 1a receptor agonist FE 202158 (10 pmol·kg(-1)·min(-1)), or vehicle (0.9% NaCl) (n = 6 each). The von Willebrand factor antigen activity relative to hemoglobin concentration (vWF:Ag/Hb ratio) was measured at different time points during the 120-min study period. Maximal vWF:Ag/Hb ratio expressed as percentage of baseline level was significantly increased compared to vehicle-infused animals (3 ± 2%) in the desmopressin (40 ± 6%, p < .001) and arginine vasopressin groups (25 ± 4%, p < .001). The ratio for the FE 202158 group was not statistically different from the sham group (9 ± 2%, p = .208). Notably, maximal vWF:Ag/Hb ratio was lower in the FE 202158 than the arginine vasopressin group (p < .005). Unlike the mixed vasopressin type 1a receptor/vasopressin type 2 receptor agonist arginine vasopressin, the selective vasopressin type 1a receptor agonist FE 202158 does not release von Willebrand factor. Because von Willebrand factor is involved in coagulatory and inflammatory pathways during septic shock, future studies should clarify the role of the vasopressin type 2 receptor-mediated von Willebrand factor increase by arginine vasopressin and the potential benefit of selective vasopressin type 1a receptor-agonists like FE 202158.