As the dimensions of concrete components continue to grow, it becomes critical to assess the validity of existing models for predicting concrete drying shrinkage. Utilizing theoretical thickness as a key parameter, a 360-d drying shrinkage test was conducted on four groups of specimens. The objective was to quantify the impact of the size effect on ultimate drying shrinkage values and associated diffusion ratios. Further, the study evaluated the applicability of current shrinkage models for specimens with varying theoretical thicknesses. The study led to adjustments in the ultimate values of established drying shrinkage models, incorporating observed size effects. Validation of these modified models was performed using previous published experimental data. Our findings revealed that ultimate drying shrinkage values are significantly influenced by the size effect, which can be described as a function of the specimen’s theoretical thickness h₀. Specifically, a 50.2 % reduction in 360-d drying shrinkage was observed when h₀ increased from 50 to 400 mm. The prediction accuracy of existing drying shrinkage models declined as the theoretical thickness increased, registering prediction errors ranging from 23.0 % to 53.3 % for h₀ > 100 mm. The modifications introduced in this study markedly enhance the prediction accuracy of existing drying shrinkage models for large-size concrete structures.