Abstract: In light of the deficiencies inherent in conventional concrete water content detection methodologies, including elevated costs and constrained detection scales, a concrete cross-section water content detection approach founded upon ultrasonic tomography imaging technology is hereby proposed. Through theoretical modelling and multi-scale tests, a quantitative relationship model was established between ultrasonic velocity, porosity, and pore water saturation (or moisture content), with a prediction error of less than 0.6%.; an ultrasonic tomography system integrating LTI ray tracing and SIRT inversion algorithms is developed, and the real-time conversion of the sound velocity cloud The map from the sound velocity cloud to the distribution cloud of water content is realised; a two-dimensional cementitious sand test specimen (with dimensions of 300 mm×300 mm×30 mm) and a three-dimensional concrete test specimen (with dimensions of 100 mm×100 mm×300 mm) are adopted to determine the water content of concrete sections. The validation of the method is demonstrated by the capillary water absorption test, which shows that it can accurately invert the distribution of cross-sectional moisture content.