The relaxation of intestinal smooth muscle by adrenaline is due to the cessation of the spontaneous discharge of action potentials which is associated with a rise in membrane potential (Biulbring, 1955;Burnstock, 1958). The magnitude of the hyperpolarization depends on the membrane potential prevailing at the moment when adrenaline is applied (Biilbring & Kuriyama, 1963). Though it is known that the membrane potential is related to the degree of stretch applied to the muscle it has not been possible in the past to set it with regularity at a certain level in different preparations, because the definition of the initial muscle length has been inadequate. Resting length cannot be determined because spontaneous activity prevents complete relaxation. In situ length depends on the spontaneous tone at the moment of excision. Thus preparations of equal in situ length, mounted at a certain length or at a certain tension, behave quite differently.The relation between muscle length and muscle tension is complicated by the active production of tension in response to stretch, which prevents the distinction between resting and active tension, the latter being an unknown fraction of the total tension recorded. On the other hand, tension is related to the cross-sectional area of the muscle. When the muscle is stretched, the deformation of the muscle fibres can be measured not only by the increase in length but also by the reduction in cross-sectional area. An estimate of this is given by the ratio of muscle weight (mg) to its length (mm). The W:L ratio takes into account the variable weight of different muscles, and thus it would be expected to provide the standard conditions required for the comparison of different preparations.In the present work reproducible conditions have been obtained by stretching the muscle to a given length in relation to its weight. Within certain limits the initial membrane potential as well as the changes produced by adrenaline were found to be proportional to the W:L ratio.