1. Acetylcholine produced end-plate current (e.p.c.) noise is shown to be the results of statistical fluctuations in the ionic conductance of voltage clamped end-plates of Rana pipiens.
2. These e.p.c. fluctuations are characterized by their e.p.c. spectra which conform to a relation predicted from a simple model of end-plate channel gating behaviour.
3. The rate constant of channel closing α is determined from e.p.c. spectra and is found to depend on membrane potential V according to the relation α = BeAV (B = 0·17 msec−1±0·04 s.e., A = 0·0058 mV−1±0·0009 s.e. at 8° C) and to vary with temperature T with a Q10 = 2·77, at −70 mV. A and B in this expression both vary with T and therefore produce a membrane potential dependent Q10 for α.
4. Nerve-evoked e.p.c.s and spontaneous miniature e.p.c.s decay exponentially in time with a rate constant which depends exponentially on V. The magnitude and voltage dependence of this decay constant is exactly that found from e.p.c. spectra for the channel closing rate α.
5. The conductance γ of a single open end-plate channel has been estimated from e.p.c. spectra and is found not to be detectibly dependent on membrane potential, temperature and mean end-plate current. γ = 0·32±0·0045 ( s.e.) × 10−10 mhos. Some variation in values for γ occurs from muscle to muscle.
6. It is concluded that the relaxation kinetics of open ACh sensitive ionic channels is the rate limiting step in the decay of synaptic current and that this channel closing has a single time constant. The relaxation rate is independent of how it is estimated (ACh produced e.p.c. fluctuations, e.p.c., m.e.p.c.), and is consistent with the hypothesis that individual ionic channels open rapidly to a specific conductance which remains constant for an exponentially distributed duration.
7. The voltage and temperature dependence of the channel closing rate constant agree with the predictions of a simple dipole-conformation change model.