Technical problems accompanying the recording of fiber pair potentials introduce certain instability in the peak-to-peak interval (rise-time, RT) of these potentials. This study aims (1) to measure the variability observed in RT of a large number of sets of consecutive potentials recorded by a single-fiber (SF) electrode and (2) to evaluate the effect of such variability on the jitter estimation. Using a SF electrode, 140 sets of consecutive potentials were recorded from the m. tibialis anterior of four healthy subjects. For each set, the rise-time variability (RTV) was calculated as the standard deviation of the RTs of the discharges within that set. The effect of RTV in the estimation of jitter from simulated fiber pairs with controlled values of neuromuscular jitter was analyzed. The RTVs of sets visually assessed as produced by a "single-fiber" were always less than 20 μs, whereas those of "composite" sets were normally higher than 20 μs. We found that the RTV always increased the estimated jitter of fiber pairs. Such increment depended on the amount of neuromuscular jitter. The RTV provides an estimate of the possible error introduced in jitter assessment. This could be important for the diagnosis of mild clinical manifestations of myasthenia gravis, myopathies, and Duchenne dystrophies.