The aim of the present study was to compare the effects of three preferential (BRL 37344, SR 58611, CL 316 243) and a partial (CGP 12177) β-adrenoceptor (β₃-AR) agonists on the contractility of ventricular strips sampled from various mammalian species including humans. In the human heart, all β₃-AR agonists tested decreased contractility by 40 to 60% below control with an order of potency: BRL 37344 > CL 316 243 = SR 58611 ≫ CGP 12177. In the dog, the negative inotropic effects produced by β₃-AR stimulation were less pronounced than in humans, ≈30% below control. The order of potency of β₃-AR agonists was CGP 12177 > BRL 37344 = SR 58611 ≫ CL 316 243; i.e., very different from that observed in humans. In rat, only BRL 37344 was efficient to decrease contractility. In guinea pig, only CL 316 243 significantly reduced peak tension. In both species, the reduction in peak tension did not exceed 20 to 30%. Finally, in the ferret, none of the agonists tested induced a negative inotropic effect. In dog, the negative inotropic effects of CGP 12177 were not modified by nadolol, but were abolished by bupranolol, a β_1–3-AR. β₃-AR transcripts were detected in the dog but not in the rat ventricle by using a reverse transcription-polymerase chain reaction assay. We conclude that cardiac negative inotropic effects related to β₃-AR agonist stimulation vary markedly depending on the species. A comparable interspecies variation previously has been reported concerning the lipolytic effects of β₃-AR agonist stimulation. Our study demonstrates that the pharmacological profile of a β₃-AR agonist on the human myocardium cannot be extrapolated from usual animal models.