This paper revisits the bicycle model for robust lateral control of autonomous front-steered vehicles to account for uncertainty in mass and inertia resulting from loading the vehicle. Unlike previous studies that tacitly assume no displacement of the center of mass due to loading, we provide a model which represents the uncertainty due to variation of loads as a positive semi-definite additive perturbation to the mass matrix. The question of how stabilizable the open-loop vehicle dynamics are, naturally follows because of additional loading. We address this problem as by determining the worst case loading distribution subject to a constraint on the maximum load added. We provide a bound on the minimum load that can be added that alter the handling characteristic of the unladen vehicle (i.e., from being an understeered vehicle to an oversteered vehicle). The purpose of lateral control is to enable the vehicle to track a specified trajectory with available information, namely, cross-track error, heading error and its rate with respect to the specified trajectory. We provide a robust fixed-structure controller that uses available feedback information and accommodates the bounded uncertainty in mass matrix. We also provide a numerical corroboration of the effectiveness of our approach.