This paper focuses on the dynamics and control challenges in manipulating structural modules with space robots for in-space assembly. First, dynamic models of the robotic assembly system are established using the Newton–Euler and Lagrange equations, where the attitude of the structural modules is represented by Euler angles. Next, motion planning in the task space is conducted using Bézier curve-based interpolation techniques to generate the parameterized reference position and attitude trajectories for the structural modules. And velocity continuity at critical path points is ensured by imposing numerical constraints on the Bézier control points. Additionally, a nonlinear model predictive control (NMPC) framework is constructed to enable structural modules to follow the planned trajectory. This method, based on the dynamic model of the system, solves a finite-time optimization problem to determine the input at the current moment. Finally, numerical simulations are carried out to verify the efficacy of the proposed planning and control algorithms. The simulation results show that the robotic system can accurately track the planned trajectories and smoothly move from the initial pose to the target pose during the assembly task.