<p>Understanding earthquake-induced cyclic liquefaction—including pore pressure generation, stiffness degradation, and their representation in constitutive models used in numerical simulations—is essential for reliable liquefaction hazard assessment. This study evaluates two widely used models: (1) the curve-fitting–based Finn–Byrne model and (2) the elastoplastic UBCSAND model. Initial calibration was performed to verify parameter consistency and identify the adjustments required for subsequent evaluation. Cyclic simple shear (CSS) test data for Nevada Sand and Fraser River Sand were used to assess each model’s ability to predict the pore water pressure ratio (<i>r</i>_<i>u</i>) and cyclic shear strain (<i>γ</i>) under constant cyclic shear stress. As expected, fitting parameters—particularly stiffness-related parameters—required modification to achieve accurate simulations. Further evaluation involved simulating five Fraser River Sand CSS tests with varying relative densities, overburden stresses, and initial shear stresses using both UBCSAND and Finn–Byrne. Additional analyses examined the sensitivity of model response to frequency content and stiffness parameters, which were manually adjusted to reproduce laboratory pore pressure build-up. The results show that UBCSAND effectively captures the full dynamic response, from initial loading through liquefaction, including flow liquefaction and cyclic mobility. The model also reproduced the flow liquefaction state (FLS) and steady state line (SSL), with a characteristic “butterfly” pattern at very low effective stresses. In contrast, the Finn–Byrne model performs well in simulating pre-liquefaction behaviour but tends to overpredict and inadequately represent the post-liquefaction response. By identifying key parameter sensitivities and model limitations, this study provides insights to support improved calibration and the future development of more robust constitutive models for liquefaction analysis.</p>