1. Confirming the results of Dudel and Kuffler (1961a), substantial increases in the tonicity of the extracellular solution around the opener of the dactyl of the walking leg produce about a four-fold increase in the frequency of miniature excitatory junctional potentials (min.e.j.p.s) (Table 1). In the frog, similar changes in tonicity produce hundred- or thousand-fold increases in frequency. 2. Increases in tonicity around the opener increase the amplitude of the min.e.j.p.s (Fig. 1). It is suggested that this effect is produced, at least in part, by changes in the input resistance of the muscle fibers produced by shrinking. 3. At the opener, large increases in tonicity produced by adding sucrose appear to produce small decreases in the amplitude of the excitatory junctional potentials (e.j.p.s) (Table 2). These solutions also decrease the excitability of the motor nerve, so some of the apparent decrease in quantal output may be owing to conduction failures in the nerve terminals. When the tonicity is increased by adding NaCl there is an increase in e.j.p. amplitude (Fig. 2). It is concluded that at this junction hypertonic solutions produce nothing like the profound decrease in stimulated release that is characteristic of the frog neuromuscular junction. 4. At the opener, ethanol (0.5–10 per cent, v/v) does not produce a detectable increase in min.e.j.p. frequency. A similar lack of effect is shown by 10−4 to 10−3 M ANS. Both agents greatly increase spontaneous release rates in the frog. 5. At the opener, the Ca2+-carrying ionophore, X-537A, produces a transient, substantial increase in min.e.j.p. frequency (Fig. 3). This shows that the failure of the other agents to increase the rate of spontaneous quantal release cannot be owing to an insufficient supply of releasable quanta (Table 3). 6. At the ventral superficial flexor of the abdomen a rise in tonicity produces a marked increase in min.e.j.p. frequencies (Fig. 4, 5). Increases in tonicity achieved by adding NaCl decrease the amplitude of the e.j.p. (Fig. 6). Both ethanol and ANS stimulate spontaneous release. 7. We suggest that, at present, unitary theories about the mechanism for transmitter release should be viewed critically.