Locating peaks of transformed anomalies provides an effective perspective to detect lateral changes in potential field data. Recently, a number of peak detection methods based on the curvature and parabola-based approaches have been introduced to interpret potential field data. As a result of differences in their calculations, the various methods produce slightly different results for any given source geometry, depth and contrast or magnetization. The purpose of this study is to compare the performance of these methods in locating peaks of transformed anomalies. To obtain optimum results, these methods have been tested on synthetic datasets and also real gravity data from the Red River Trough, Vietnam. The results show that the parabola-based method and its modified version are less effective for detecting peaks in the intersection regions of sources. In addition, although the curvature-based method does not have such a problem, it can bring the secondary maximum locations.