Despite its potential, daylighting strategies in school classrooms in the tropical climate regions is little explored in the literature. The use of two-sided or bilateral daylight opening, as well as the self-shading mechanism using sloped walls, are currently seen as potential strategies to achieve good daylighting in tropical buildings. This study thus aims at exploring and optimizing self-shading design possibilities for creating daylight-friendly, tropical elementary school classrooms with bilateral openings, by means of sensitivity and uncertainty analyses. Design parameters such as façade orientation, window-to-wall ratio, window elevations, wall slopes, interior surface reflectance and glazing transmittance are considered in the model of a hypothetical, typical classroom without shading devices. Climate-based daylight metrics such as UDI250-750 lx, sDA300/50% and ASE1000,250 are utilized as the performance indicators. Computational modelling with Honeybee [+] in Rhinoceros/ Grasshopper platform is employed to conduct the annual daylight simulation. Multi-objective optimizations using genetic algorithms (GA) through Octopus with SPEA-2 algorithms are performed to determine the optimum design solutions following the sensitivity and uncertainty analysis. The slope and WWR of the wall that faces west, southwest, south, or southeast, has the strongest influence on the defined objective function that includes all daylight metrics. To achieve the optimum daylight performance, design of the opposing window façades of school classrooms with bilateral openings should not be identical.