In the preceding paper, we described the parallel effect of amphipathic drugs on the fluidity and adenylate cyclase activity of pigeon erythrocyte membrane. This parallelism was found when the cyclase activity was assayed in the presence of a guanyl nucleotide and with Mg-ATP as the substrate after a preincubation of the membrane with the drugs [Salesse, R., Garnier, J., Leterrier, F., Daveloose, D., & Viret, J. (1982) Biochemistry (preceding paper in this tissue)]. However, when the regulatory protein (N) and the catalytic unit (E) were precoupled by GppNHp or fluoride before the action of the drugs, the cyclic AMP production was never inhibited. Thus, the drug-induced fluidization appeared to interfere with the efficiency of the activating coupling between N and E. Chlorpromazine even enhanced the cyclase activity: if the catecholamine receptor (R) repressed the cyclase activity in the absence of hormone [Rodbell, M. (1980) Nature (London) 284, 17-22], the loss of R molecules with chlorpromazine would prevent this inhibition and lead to hyperactivity of the enzyme. On the other hand, the comparison between two states of the adenylate cyclase system, (1) N and E reversibly precoupled in the presence of GTP and (2) R, N, and E precoupled in the presence of GTP plus isoproterenol, showed no difference between the activity curves at various drug concentrations: this may be interpreted as a permanent coupling of R and N. The main control exerted by fluidity on the activity of the adenylate cyclase system would thus be at the level of the activating coupling between the N subunit and the catalytic unit in pigeon erythrocyte membrane.