1. End-plate potentials (e.p.p.s) were recorded with a surface electrode from frog neuromuscular junctions blocked with high Mg and low Ca to study post-tetanic potentiation(potentiation). 2. Potentiation is found to decay exponentially over most of its time course. 3. The time constant tsu(H) characterizing this exponential decay is a function of the previous history (frequency and duration) of stimulation. For example, tsu(H) increased from about 20 sec following a few impulses to over 70 sec following more than 1000 impulses. 4. A new method is presented to obtain estimates of the rise of potentiation (uncontaminated by facilitation or an intermediate facilitatory process) during repetitive stimulation. It is found that potentiation is present following short trains of impulses and continually increases in magnitude with the duration of the conditioning stimulation. Potentiation was at a maximum immediately following the conditioning trains. 5. The relationship between P(T), the magnitude of potentiation immediately following repetitive stimulation, and tsu(H), the time constant for the decay of this potentiation, is given by tsu(H)=Ae-P(T)/B, where A=19-8 plus or minus 5-1 sec (mean plus or minus S.D. of an observation) and B increases from 2.2plus or minus 2-1 to 5-7 plus or minus 2-7 as the stimulation rate increases from 5 to 30/sec. 6. The value of A in the above equation can be considered to represent the minimal time constant for the decay of potentiation--that is, the time constant for decay after a simgle impulse. 7. Evidence is presented for a afacilitatory process with a time constant of decay of about 3 sec which is intermediate in duration between facilitation and potentiation. 8. It is suggested that repetitive stimulation has a dual effect on potentiation; each impulse (1) adds an increment of potentiation and (2) increases tsu(H), the time constant for the decay of potentiation.