A preliminary estimation of construction costs is a crucial aspect of any civil engineering project. This paper introduces a framework for estimating construction costs that ensures both accuracy and explainability. The approach utilizes six machine learning models and Shapley additive explanations.