Piggy-backed anchors present a promising solution to address the limitations of single Drag Embedment Anchors (DEAs). A series of centrifugal model tests were performed on the installation of piggy-backed anchors with varying spacings and attachment points in saturated sand using a magnetometer system. The experimental results validated the proposed theoretical prediction model, which can identify the trajectory, capacity, orientation, and soil failure wedge slip surface. Key findings include: (1). the embedment depth of the second anchor in piggy-backed anchors is significantly greater than that of a single anchor, resulting in the total capacity of piggy-backed anchors exceeding twice that of a single anchor at appropriate spacing; (2). investigating spacing effects shows that capacities are less than twice that of the single anchors when spacing is less than 1Lf due to overlapping failure soil wedges; optimal spacing was found to be 2Lf to 3Lf; (3). piggy-backed anchors with attachment points at the back of fluke exhibit better embedment performance and capacity stability compared to those with attachment points at the padeye. Finally, the prediction model explored the capacity performance of piggy-backed anchors at different loading angles at the final embedment depth, revealing a two-stage failure process compared to single drag anchors.