Assembly accuracy is crucial to the working performance of precision machinery, which is affected by the surface profile errors of parts in the manufacturing process and deformation under the action of assembly force. Hence, this study proposes an assembly accuracy analysis method that integrates the surface profile error of parts and deformation under force. Firstly, a small-displacement torsor model is used to model the pose error and a shape error model is constructed by combining the basis function superposition method, and a comprehensive surface error model is established by introducing a Gaussian function to generate random noise. Subsequently, a finite element method is used to analyze deformation of parts under force conditions and its influence on assembly pose, and non-uniform rational B-splines (NURBS) are used to fuse and reconstruct the surface shape and part deformation, generating a surface model of the part that simultaneously considers the surface profile error and force deformation. Finally, based on this model, the fit state between surfaces is evaluated by combining the nearest projected point method, and the assembly accuracy is then calculated by an optimization algorithm. The results of 1000 simulation experiments on planar bolted joints and interference fits of column surface show that different assembly forces or overloads affect the final assembly accuracy; the proposed method is able to carry out effective evaluation and provide guidance for actual assembly.