1. The kinetics of tubocurarine inhibition were studied at the post-synaptic membrane of frog skeletal muscle fibres. Acetylcholine (ACh) and (+)-tubocurarine were ionophoresed from twin-barrel micropipettes, and the membrane potential of the muscle fibre was recorded intracellularly. Tubocurarine-receptor binding was measured by decreases in the response to identical pulses of ACh. 2. The responses to both ACh and tubocurarine had brief latencies and reached their maxima rapidly. It is suggested that under these conditions the kinetics of tubocurarine action are not slowed by diffusion in the space outside the synaptic cleft. 3. After a pulse of tubocurarine, recovery from inhibition proceeds along a roughly exponential time course with a rate constant, 1/tau off approximately equal to 0.5 sec-1. This rate constant does not depend on the maximal level of inhibition and varies only slightly with temperature (Q10 = 1.25). 4. After a sudden maintained increase in tubocurarine release, the ACh responses decrease and eventually reach a new steady-state level. Inhibition develops exponentially with time and the apparent rate constant, 1/tau on, is greater than 1/tau off. When the steady-state inhibition reduces the ACh response to 1/n of its original level, the data are summarized by the relation, 1/tau on = n(1/tau off). 5. When the ACh sensitivity is reduced with cobra toxin, both 1/tau on and 1/tau off increase. Thus, the kinetics of tubocurarine inhibition depend on the density of ACh receptors in the synaptic cleft. 6. After treatment with collagenase, part of the nerve terminal is displaced and the post-synaptic membrane is exposed directly to the external solution. Under these circumstances, 1/tau off increases more than tenfold. 7. Bath-applied tubocurarine competitively inhibits the responses to brief ionophoretic ACh pulses with an apparent equilibrium dissociation constant, K = 0.5 microM. 8. In denervated frog muscle fibres, extrasynaptic receptors have a lower apparent affinity for tubocurarine. After a pulse of tubocurarine, inhibition decays tenfold more rapidly at these extrasynaptic sites than at the synapse. 9. It is suggested that each tubocurarine molecule binds repeatedly to several ACh receptors before escaping from the synaptic from the synaptic cleft and that the probability of this repetitive binding is enhanced because the nerve terminal presents a physical barrier to diffusion out of the cleft. Consequently, the receptor transiently buffer the concentration of tubocurarine in the cleft, and the macroscopic kinetics of inhibition are much slower than the molecular binding rates for tubocurarine.