1 The role of calmodulin (CaM) in transmitter release was investigated using liposomes to deliver CaM and monoclonal antibodies against CaM (antiCaM) directly into the frog motor nerve terminal. 2 Miniature endplate potentials (MEPPs) were recorded in a high K + solution, and e ects on transmitter release were monitored using estimates of the quantal release parameters m (number of quanta released), n (number of functional transmitter release sites), p (mean probability of release), and var s p (spatial variance in p). 3 Administration of CaM, but not heat-inactivated CaM, encapsulated in liposomes (1000 units ml 71 ) produced an increase in m (25%) that was due to an increase in n. MEPP amplitude was not altered by CaM. 4 Administration of antiCaM, but not heat-inactivated antiCaM, in liposomes (50 ml ml 71 ) produced a progressive decrease in m (40%) that was associated with decreases in n and p. MEPP amplitude was decreased (15%) after a 25 min lag time, suggesting a separation in time between the decreases in quantal release and quantal size. 5 Bath application of the membrane-permeable CaM antagonist W7 (28 mM) produced a gradual decrease in m (25%) that was associated with a decrease in n. W7 also produced a decrease in MEPP amplitude that paralleled the decrease in m. The decreases in MEPP size and m produced by W7 were both reversed by addition of CaM. 6 Our results suggest that CaM increases transmitter release by mobilizing synaptic vesicles at the frog motor nerve terminal.