Abstract: We present a novel cascade tire-force estimation framework based on the sliding-mode observer method. We estimate each longitudinal tire force on the basis of the single-wheel vehicle model, which uses the angular velocity and driving torque of the wheel as inputs. Then, we apply a simplified 2-degrees-of-freedom (2-DOF) vehicle dynamic model to design the front and rear lateral-tire-force sliding-mode observer, and use the estimated longitudinal tire forces, the front-wheel angle, the lateral acceleration, and the yaw rate as the observer inputs. We experimentally evaulate the proposed tire-force estimation method on the high-precision vehicle dynamic software, veDYNA, and compare the estimation results with those of the extended Kalman filter (EKF) method. The results demonstrate that the sliding-mode observer has relatively higher accuracy in estimating tire forces.