The concept of a seismic cycle, where the stress on a fault repeatedly builds up over a long period of time and then is rapidly released in a large earthquake, influences studies of both the basic physics of faulting and applied research aimed at estimating earthquake hazards. This hypothesis suggests that large earthquakes might be quasi periodic and that the probability of a particular portion of a fault rupturing twice in quick succession should be low. However, this basic hypothesis has been difficult to verify owing to the long repeat times of the largest earthquakes on most faults. East Pacific Rise (EPR) transform faults are an advantageous location to evaluate the seismic cycle hypothesis owing to their fast slip rates and the moderate size (∼Mw 6) of their largest earthquakes. Using surface-wave based determinations of the relative separations between earthquake centroids, I document 16 pairs of Mw≥5.5 events that had overlapping ruptures. The distribution of interevent times for these pairs is tightly clustered around 5 yr (with a coefficient of variation ∼0.2) indicating that quasi periodicity may be prevalent for the largest events on these faults. Moreover, I find no pairs of overlapping Mw 5.5–6.2 earthquakes separated by less than 50 cm of elapsed plate motion, indicating that the two basic features of the seismic cycle hypothesis are evident in the timing of large EPR transform mainshocks. I have also confirmed earlier results demonstrating a high degree of short-term predictability of EPR mainshocks by combining teleseismic and hydroacoustic earthquake catalogs. Thus, there appears to be a high degree of both short and long-term predictability on EPR transforms.