High‐resolution inverse synthetic aperture radar (ISAR) imaging can produce electromagnetic images of a target in the range and cross‐range domain. However, to make full use of an ISAR image in target detection and recognition, we need to rescale the image in a homogeneous range and cross‐range domain, and further, to estimate the effective rotational velocity (RV) of the target. A precise ISAR image cross‐range scaling method is proposed. A candidate RV is obtained based on the bisection method. Then, the effective rotation angle (RA) between two sequential rescaled images is characterised by the RA between the major axes of the images. These axes are extracted by principle component analysis (PCA). A precise estimation of the RV is obtained through iterative PCA and bisection. Simulated and real data experimental results validate the effectiveness of the proposed method.