Negative shear stress values often appear in the falling section of a bond-slip experimental curve of a fiber-reinforced polymer (FRP) sheet–concrete interface; the reason for this phenomenon is unclear. Defining this phenomenon as the “shear stress reverse phenomenon,” we attempt to explore its mechanism and influence on the bond-slip behavior and bond-slip model through theoretical and numerical analyses, and experimental verification. The theoretical analysis revealed that an “interface mismatch” often occurs during the shear failure of an FRP sheet–concrete interface, resulting in the formation of discontinuous microcracks at the interface, which cause fluctuations and make the interface slip field discontinuous. Owing to the quadratic derivative relationship between the slip and interfacial shear stress, the fluctuations in the slip curve lead to the shear stress reverse phenomenon. The validity of this theoretical concept was verified by analyzing the experimental interfacial bond-slip curves obtained by scholars in previous studies, and by performing double shear experiments on three fully-bonded FRP sheet–concrete interface specimens. The experimental results confirmed the existence of fluctuations in the interfacial slip field and the shear stress reverse phenomenon in the bond-slip curve. Finally, the failure process of the FRP sheet–concrete interface was analyzed through the finite element method, and the fluctuation of the slip and existence of the shear stress reverse phenomenon at the interface were observed in the calculation results. The results revealed that the shear stress reverse phenomenon is caused by special displacement boundary conditions, rather than the material properties. Therefore, while establishing a bond-slip model by fitting experimental curves, bond-slip curves that exhibit the shear stress reverse phenomenon should be deleted in their entirety, instead of only deleting the negative shear stress data in such curves.