By conducting a series of shear-torque-controlled (STC) and shear-displacement-controlled (SDC) ring shear tests under undrained conditions, the effects of cyclic loading frequency, shear displacement rate, and overconsolidation ratio (OCR) on the pore-pressure generation were examined and analyzed by means of an energy approach. Cyclic STC tests demonstrated that as the frequency of loading increased, the shear energy (W-total) as well as the total shear displacement (l(total)) until liquefaction substantially decreased, while the number of cycles to liquefaction (N) increased with frequency. Nevertheless, SDC tests showed that W-total, l(total), N did not vary with frequency. This dependency of W-total on the loading frequency in STC tests was inferred to be due to the different resultant shear displacement amplitude after shear failure but before liquefaction during cyclic shearing. The results of STC tests on samples with different OCRs showed that all these three parameters of W-total, l(total), N increased with OCR. The SDC tests at different shear-displacement amplitude (Delta l(max)) showed that there existed an optimal Delta l(max) at which W-total was minimum. Delta l(max) smaller than this optimal value was probably not as effective at enforcing the grains to adjust their position, and then it was difficult for the volume shrinkage to occur and pore-water pressure to generate; while an increase in Delta l(max) from this optimal value led to extra energy consumption probably due to grain crushing and heat transferring from grains friction, and then elevate the value of W-total for liquefaction. These results proved that pore-pressure generation in undrained cyclic loading was strongly dependent on the shear-displacement amplitude during the shearing.