In the layer-based additive manufacturing (AM) processes, a three-dimensional (3D) model is converted into a set of two-dimensional (2D) layers. Due to such conversion, one of the major problems in the layer-based AM processes is the poor surface finish associated with the layer-based stair-stepping effect. However, the surface finish is critical for various microscale applications such as micro-optics and microfluidics. The adoption of AM technologies as a means for fabricating end-use microcomponents and tooling has been limited by such poor surface finish. The aim of this research work is to apply the state-of-the-art meniscus approach and controlled cure depth planning in the mask image projection-based microstereolithography (MIP-μSL) process to address its surface finish challenge. Mathematical models of meniscus shapes and cure depths are developed for the MIP-μSL process. Related process parameters including the minimum meniscus points, sliced layer shapes for forming meniscus, grayscale image values, and Z offsetting values are optimized to achieve the minimum approximation errors between a built part and a given nominal geometric model. A set of test cases with various curved surfaces are designed to test the developed smooth surface fabrication method. The experimental results verify the effectiveness of the proposed methods for the MIP-μSL process.