A thin rectangular steel wall in a steel shear wall structure always simultaneously sustains the lateral load and the gravity load. The gravity load can affect the shear strength of a steel shear wall. However, this effect is not considered in most of the research and standards, which may lead to potential danger in practice. From the previous study of the authors, the shear strength reduction was not only influenced by the load magnitude but also by the vertical stress distribution. For a simply-supported thick square wall, i.e. width to thickness ratio smaller than 100, the stress distribution can be accurately described in a cosine form. However, for a thin wall under compression and in-plane bending, the cosine distribution will largely overestimate the vertical stress, especially when the walls enter the post-buckling condition. To narrow the knowledge gap, this paper proposed a vertical stress distribution in a three-segment form, i.e. in both edge-segments, a combination of linear and cosine functions from the edge stresses to the minimum stress, while in the middle segment, the stress distribution is constant and equal to the minimum stress. Two strategies, i.e. effective width method and Bedair’s method, are chosen to determine the width of the edge portion. A finite element model (FEM) is developed to evaluate the proposed distribution. The FEM has been verified using the results of our own experiments and tests done by Zaraś et al. The results show that the proposed three-segment stress distribution can well describe the behavior of thin walls of different slendernesses and stress gradients. The cosine distribution obtained from theoretical solution and the effective width model by Bedair are also discussed.