We wish to report the chiral synthesis and biological evaluation of 3 (CL 316,243), a benzodioxole-containing phenethanolamine, which is a highly potent stimulant of rat adipocyte lipolysis (β3 effect) but an extremely poor β1 and β2 agonist, with β3 selectivity greater than any previously reported agent. In vitro, it had no effect on guinea pig atrial contraction rate (β1 effect) and only limited ability to inhibit insulin-stimulated [14C]glucose incorporation into glycogen in isolated rat soleus muscle (β2 effect). Binding experiments showed it had much lower affinity for β1 (rat heart membranes) and β2 (rat soleus muscle membranes) receptors than reference compounds. In obese (ob/ob) mice (a non-insulin-dependent diabetes model), 7-week treatment reduced weight gain (from 13.0 ± 0.6 g to 7.1 ± 0.6 g, p < 0.01) despite slightly increased food intake, and normalized hyperglycemia to euglycemia (maintained throughout the experiment). Functional in vitro assays and binding experiments confirmed it is a highly potent β3 adrenoceptor agonist with very low β1/β2 affinity. Given its potent antiobesity and antihyperglycemic effects in animal models, it offers a possibility for treating non-insulin-dependent diabetes and obesity without undesirable β-mediated side effects. The compound is now in Phase I clinical trials.