Abstract: Objective: To investigate whether water-equivalent diameter (WED) modulates the dose-efficiency gain of Freeze motion correction in low-tube-voltage coronary computed tomography angiography (CCTA), and to compare the performance of different generations of Freeze algorithms. Methods: In this prospective study, 104 participants who underwent low-tube-voltage CCTA were enrolled. Images were reconstructed from the same raw scan data for Freeze 0, Freeze 1.0, and Freeze 2.0. The dose-normalized contrast-to-noise ratio (nCNR) and related objective metrics were derived from standardized region-of-interest measurements, and the subjective image quality scores and diagnostic usability were assessed. Correlation analyses were performed to evaluate the associations of WED with the baseline nCNR and gain metric ΔnCNR. A linear mixed-effects model was used to test the freeze × WED interaction, and segmented linear regression was applied to explore a practical WED threshold. Results: WED was significantly and inversely correlated with baseline nCNR, indicating a reduced baseline dose-normalized imaging efficiency with increasing attenuation burden. WED was also negatively correlated with ΔnCNR, suggesting that the dose-efficiency gain from Freeze motion correction decreased as WED increased. The mixed-effects model confirmed significant Freeze × WED interactions. Overall, generational Freeze upgrades improved image quality per unit dose, with nCNR increasing from 5.04±1.28 (Freeze 0) to 6.03±1.51 (Freeze 2.0) (all P < 0.001), accompanied by parallel improvements in subjective scores and diagnostic usability. Segmented regression and bootstrap analyses suggested that a WED of approximately 28.0 cm may represent an exploratory threshold beyond which the marginal benefit of motion correction is attenuated. Conclusion: WED significantly modulated the dose-efficiency gain of freeze-motion correction in low-tube-voltage CCTA. Higher WED values were associated with reduced baseline imaging efficiency and diminished marginal benefit from motion correction, thereby providing an important reference for individualized optimization.