Abstract: Abstarct: Prestressed cables are the main load-bearing members of long-span bridges. In the accident of a vehicle fire on the bridge, if the cables suffer fire damage, they are difficult or impossible to replace, which could also jeopardize the safety of the entire bridge. Existing fire resistance tests for cables predominantly simulate non-load conditions. Furthermore, both domestic and international standards have yet to establish methods for fire resistance testing of cables under loading. A method for fire resistance testing of prestressed cable surrogate models under loading is proposed in this article, with a focus on two aspects: the experimental model and the equipment. The cable model is equivalently represented as a surrogate model composed of heat transfer steel wires and load-bearing steel wires, which ensures the replacement of small loading for large-scale prestressed cable models under fire conditions while also reproducing the non-uniform temperature rise across the cable cross-section during experiments. Additionally, quantification method for the cross-sectional design of a surrogate model is proposed. Using the temperature of the steel wires on the external surface of the cable as a control parameter, recommendations for the minimum dimensions of the cable surrogate model in the experiment are provided. A coaxial changing-diameter force transmission system and a mobile fireproof enclosure are designed to meet the functional requirements of non-full-length fire testing for prestressed cable models of varying diameters. Based on this, the implementation process for fire resistance testing of the load-bearing cable models, the heating system, and the conditions for test termination are provided in this article. Verification tests for the fire resistance of the load-bearing cable models were conducted. Strain monitoring data indicated that the surrogate models were uniformly loaded before the occurrence of the fire, and they achieved the working prestress state of the cables. During the temperature rise of the fire, all load-bearing steel wires experienced fire creep fracture, and the model cross-section exhibited a clear temperature gradient. In comparison to non-load-bearing fire resistance tests, the load-bearing fire conditions can lead to the opening of fire protection seams, thereby accelerating the temperature rise and fire failure of the test model. This confirms the necessity and rationality of the experimental method presented.