In order to reveal the influence law of the stud spacing on the behavior of steel-UHPC lightweight composite deck structures, variable amplitude fatigue tests were accomplished for three LWCD beam specimens and the stud spacing was chosen as the core parameter in the tests (i.e., 100 mm, 150 mm, 300 mm). In the fatigue tests, the influence of the stud spacing on the fatigue performance of the LWCD was revealed and the combined tension-shear fatigue properties of the steel deck plate at stud root positions were obtained. In the residual capacity test after the fatigue tests, the influence of stud spacing on the cracking evolution mechanism in UHPC and on the residual capacity of the specimens was captured. The fatigue tests indicated that when the stud spacing was 300 mm, the unit-load induced slips at the steel-UHPC interface were much higher than those in the other two specimens, but the slips did not increase significantly during the fatigue testing process. When the stud spacing was 100 mm or 150 mm, the strain at the bottom flange of the U rib did not exhibit significant change, and the strains increased slightly when the stud spacing was 300 mm. On the basis of the calculation method prescribed in the code of “Specifications for Design and Construction of Highway Steel-concrete Composite Bridge” in China, the combined tension-shear fatigue state of the steel deck plate at stud root positions was analyzed, and the results implied that this method could provide conservative results. In addition, the residual capacity test after the fatigue testing showed that the smaller the stud spacing, the smaller the interfacial slip and the higher the residual flexural capacity. On the basis of the elastic-to-plastic theory and plastic theory, the residual capacities of the specimens were predicted. It was revealed that although the specimens experienced a series of fatigue loading, the actual residual capacities obtained in the tests were still higher than the predicted values, and the plastic theory exhibits a better agreement relative to the test results.