As a critical candidate technology for 5G-advanced and 6G, reconfigurable intelligent surfaces (RIS) have received extensive atten-tion from academia and industry. RIS has the promising features of passiveness, reconfigurable ability, and low cost. RIS channel estimation faces the challenges of high matrix dimension, passive estimation, and spatial-wideband effect. In this article, we analyze the impact of the spatial-wideband effect on the RIS channel to account for the propagation delay across RIS elements and estimate sparse channel parameters such as angle and gain through a super-resolution compressive sensing (CS) algorithm. The simulation results explore the influence of the spatial-wideband effect on the RIS channel and verify the effectiveness of the proposed algorithm.