<p>Seismic resilience assessment of transportation infrastructure is a critical aspect of enhancing the disaster adaptation capacity of lifeline systems. As a key performance-based indicator, seismic resilience represents the next-generation design concept of performance-based seismic design. This study presents a quantitative framework for assessing the resilience of urban and major roads affected by earthquake-induced landslides. A detailed two-dimensional numerical model was developed in PLAXIS^2D, incorporating eight slope geometries defined by two slope heights (30 m and 45 m) and four slope angles (15°, 25°, 35°, and 45°). The seismic response of slopes was analyzed under 300 recorded ground motions. Seismic fragility curves for roads were generated using permanent displacement to capture the influence of the peak ground acceleration (PGA). By integrating fragility functions with restoration models, resilience curves were derived to quantify the ability of roads to withstand seismic damage and recover functionality. The resilience index curves were developed and evaluated. The findings show that both slope geometry and PGA exert significant influence on road functionality and recovery performance. The result indicates that at lower slope (e.g. 30 m) the functionality of roads recovers faster than those of higher slope (e.g. 45 m) around 7% after three days at PGA of 0.5 g. Additionally, under the increasing of PGA (e.g. from 0.5 g to 0.7 g), the functionality decreases by around 11% for urban road, and 5% for major road at one slope geometry. Furthermore, the proposed framework provides a quantitative methodology for resilience assessment and offers theoretical support for resilience design, optimization, and post-earthquake recovery planning of transportation infrastructure.</p>