Diaphragms of young mice were superfused with Bretag´s solution that contained Cd2+ in order to block Ca2+-channels without lowering the intracellular Ca2+-concentration. A perfused macropatch pipette elicited quantal releases by depolarisation pulses and recorded synaptic currents. Depolarisation/release relations were measured with control 2 mM Ca2+ or <0.1 mM Ca2+ in the perfusate; the latter allowing negligible Ca2+-inflow. Compared to the controls, with <0.1 mM Ca2+ the threshold for release shifted to larger depolarisations, and close to the saturation level average release reached about a tenth of the controls. The time courses of release shortened slightly with increasing depolarisation, but did so with and without Ca2+- inflow. For large depolarisations that exceeded the level of release by action potentials in 2 mM Ca2+, the time courses of release were shorter without Ca2+-inflow than in 2 mM Ca2+. In order to control that the terminals really were bathed in <0.1 mM Ca2+ solution, the pressure of the perfusate at the tip of the pipette was arranged to force about half of the perfusate through the hole at the tip into the bath and the muscle, while Cd 2+ was omitted from the superfusate of the bath. Also under these conditions the depolarisation/release relations were very similar to those obtained with Cd2+ in the bath. - It is concluded that a non-Ca2+ factor of the release machinery that increases with the amplitude of depolarisation largely controls the time course of release. The inflow of Ca2+ multiplies the amount of release.