Closed loop manufacturing systems (CLMSs) with recirculating material handling devices are extensively used in various industrial environments. The performance of such systems is impacted by many factors such as the total capacity of pallets, the actual number of pallets in the system, the machine reliability and processing time, the pallet index speed, and the positions of loading/unloading points. These factors make the accurate analysis and optimization of complex CLMSs very difficult and challenging. This paper presents a new parameter coupling technique to model and analyze a wide range of CLMSs. It is an enhancement based on the existing open production line analysis with unreliable assembly machines and finite buffers. Virtual assembly machines are introduced to represent the specific phenomena of CLMSs such as the recirculation of empty pallets and the sharing of conveyor space. Two types of parameter coupling patterns, the machine parameter coupling, and the buffer capacity coupling, are introduced to reflect the characteristics of the CLMSs. The parameter coupling technique is simple and effective for analyzing a broad range of CLMSs. Comparisons between this analytic method and simulation experiments demonstrate that the proposed parameter coupling technique is fast, accurate and robust.