The subclass of beta-adrenergic receptors mediating glycogenolysis in slices from cerebral cortex of the mouse, incubated in the presence of [3H]glucose, was identified by comparing the relative potencies of agonists and the inhibition constants of antagonists to those found on reference systems. (+/-)Isoprenaline, (-)adrenaline and (-)noradrenaline produced a concentration-related glycogenolysis with Kact values of 2.2 x 10(-8) M, 2.8 x 10(-7) M and 3.6 x 10(-7) M, respectively. Zinterol, a selective beta 2-adrenergic agonist, did not produce any glycogenolytic response even in a large concentration. Salbutamol, a predominantly beta 2-adrenergic receptor agonist, elicited in a very large concentration (10(-4) M) less than 40% glycogenolysis, an effect which was not related to stimulation of beta-adrenergic receptors. The predominantly beta 1-adrenergic receptor antagonists, practolol and metoprolol shifted the concentration-response curve to noradrenaline to the right, with apparent Ki values of 8.0 x 10(-7) M and 7.6 x 10(-8) M, respectively, close to those reported in the rat heart. These various data indicate that the glycogenolytic response is selectively mediated by beta 1-adrenergic receptors. Under experimental conditions which were strictly identical to those used to measure glycogenolysis, a saturable binding of [3H]dihydroalprenolol to the slices occurred with Kd and Bmax values consistent with corresponding values on cell free preparations. Whereas the Ki values of antagonists were similar on the two systems, the Kact values of agonists on glycogenolysis were 10 times less than the Ki values for the binding of [3H]dihydroalprenolol. This suggests that the maximal glycogenolytic response is elicited for a partial receptor occupancy.