Abstract: Ultra-High-Performance Concrete （UHPC） has been widely applied in demanding engineering fields, including buildings, bridges, and tunnels, owing to its superior mechanical properties and durability. However, due to its low water-to-binder ratio and high viscosity, internal inhomogeneities, such as dry powder agglomeration and steel fiber clustering, are prone to occur during the mixing and casting processes, thereby adversely affecting the mechanical performance and long-term durability of the material. Conventional concrete testing methods primarily focus on mechanical properties and compactness, whereas studies concerning the evaluation of the internal homogeneity of UHPC remain limited. To address this issue, a vibration frequency–based non-destructive testing method for evaluating the internal homogeneity of UHPC is proposed in this study. The UHPC specimens were subjected to controlled impact excitation, and the vibration response signals were acquired using accelerometer sensors at a fixed sampling frequency. Wave velocity statistics and the Pearson correlation coefficient were employed to quantitatively characterize waveform similarity for the identification of internal homogeneity. Three types of specimens （Specimens 1, 2, and 3） were designed to represent typical conditions, namely uniform distribution, fiber-deficient regions, and steel fiber-enriched regions, respectively. Comparative experiments were conducted using both the proposed vibration frequency method and CT scanning. The results indicate that the proposed method can effectively identify internal inhomogeneities caused by dry powder agglomeration and steel fiber clustering, and the obtained results are in good agreement with the CT scanning images. Furthermore, the method can accurately characterize variations in the internal homogeneity of UHPC during production. When combined with CT scanning, the proposed approach provides a scientific basis for improving the production quality and long-term service performance of UHPC.