Abstract: Downhole liquid-phase discharge plasma rock breaking and wire explosion are the two primary implementations of high-voltage pulsed discharge technology for rock fragmentation. The aim of this study is to systematically compare the rock fragmentation performance and energy dissipation between the two methods. Homogeneous cubic cement mortar specimens, each with a side length of 200 mm and a prefabricated central borehole measuring 20 mm in diameter and 120 mm in depth, were used to simulate soft rock in laboratory experiments. A test system was constructed, consisting of a high-voltage pulse discharge device, a discharge-parameter measurement module, and a high-speed photography module. The high-voltage pulse discharge device features a 250 μF capacitor with an adjustable charging voltage range of 0-100 kV, which was used to apply high-voltage pulses to the specimens. Concurrently, the discharge-parameter measurement module recorded current and voltage waveforms, enabling comparison of the differences in energy transfer processes between the two methods. Simultaneously, the high-speed photography module captured the dynamic failure process of the specimens under shock wave loading, facilitating analysis of differences in crack propagation. In the tests, needle-needle electrodes with a fixed electrode spacing of 10 mm were inserted into boreholes and served as the discharge electrodes for generating high-voltage pulses. Four charging voltages of 7 kV, 8 kV, 9 kV, and 10 kV were used as experimental variables to compare the characteristics of soft rock fragmentation between downhole liquid-phase discharge plasma and wire explosion. High-speed photography was employed to analyze and compare crack propagation behaviors between the two methods. Quantitative assessments of surface crack density and fractal dimension were conducted to characterize crack morphology differences in the processed specimens. Statistical analysis of fragment size distribution and specific surface area further revealed distinctive fragmentation characteristics. Additionally, on the basis of the dynamic processes of energy release and transfer in rock fragmentation via high-voltage pulse discharge and the relationship between newly created surface area and fragmentation energy in fracture mechanics, time-history current and voltage data were utilized to quantify the downhole deposited energy distribution and fragmentation energy allocation efficiency. The results show that, for rock fragmentation by liquid-phase discharge plasma, the surface crack density and newly created surface area increase monotonically with the increase of charging voltage, while the average fragment size presents a monotonically decreasing trend. For rock fragmentation by wire explosion, however, the surface crack density and newly created surface area increase first and then decrease, with the average fragment size showing an inverse trend. The fractal dimension of specimen surface cracks increases with the rising voltage for both methods, with wire explosion generating a more complex crack network. As charging voltage increases from 7 kV to 10 kV, the energy deposition efficiency for rock fragmentation by liquid-phase discharge plasma increases from 55.37% to 69.43%, whereas for rock fragmentation by wire explosion, it rises from 59.78% at 7 kV to a peak of 67.47% at 9 kV before sharply dropping to 52.96% at 10 kV. Meanwhile, the fraction of rock fragmentation energy within deposited energy declines from 7.53% to 4.77% for liquid-phase discharge plasma, which accounts for 3.32%–4.17% of the total energy stored in the capacitor. For wire explosion, the fraction declines from 8.08% to 6.62%. The corresponding value accounts for 2.98%–4.83% of the total energy stored in the capacitor. This demonstrates that the wire explosion technique delivers a higher conversion efficiency from deposited energy to rock fragmentation energy. At 10 kV, the insufficient deposited energy generated by wire explosion directly degrades its rock-breaking performance at this voltage, which deviates from the consistent upward trend of rock fragmentation efficacy observed in liquid-phase discharge plasma as the applied voltage rises. The findings offer valuable guidance for the method selection and parameter optimization of high-voltage pulse discharge for rock fragmentation.