Shaft misalignment is a prevalent fault in rotating machinery, leading to increased vibration, accelerated wear, and potential failure. Early and accurate diagnosis is crucial for ensuring system reliability and safety. This work presents a novel misalignment diagnosis method based on correlation analysis of vibration signal cycles. The approach consists of order tracking, filtering, signal segmentation, and correlation mapping. Order tracking transforms the signal into the angular domain to eliminate speed variations, while a band-pass filter isolates low-order frequency components linked to misalignment. The segmented signal cycles are then analyzed for correlation, and the resulting correlation map serves as a diagnostic tool. Unlike machine learning-based methods, the proposed approach does not require historical training data. Moreover, it relies on vibration data from a single sensor, making it more practical than orbit-based techniques that require multiple sensors. The method’s effectiveness is validated through experiments on a public dataset, demonstrating its potential for robust and data-efficient misalignment detection.