We focus on the original equipment manufacturers (OEMs) who sell both new and refurbished products directly to consumers, such as a computer maker who sells products online. New products are produced according to a make-to-order policy while returned products will be refurbished and ready to ship to consumers immediately according to a make-to-stock policy. The open queueing network combines the demand functions of new and refurbished products with a generalized open queueing network that simulates the closed-loop supply chain including manufacturing, sales, return, evaluation, refurbishing and resale processes. A mathematical model is developed to find the optimal price and quantity of refurbished products. The results show the effects of several parameters on the optimal solutions;Examination of the optimality conditions reveals the different situations in which it is optimal to refurbish none, some, or all of the returned products. Refurbishing operations may increase profit or may be required to relieve a manufacturing capacity bottleneck. A numerical study identifies characteristics of the new product market and refurbished products that encourage refurbishing and some situations in which small changes in the refurbishing cost and quality provoke large changes in the optimal policy. Sensitivity studies reveal how the variability of times to manufacture and refurbish products affects the optimal policy. Categorizing returned products improves total profit by reducing the variability of refurbishing time. The study also illustrates the improvement of total profit when products are designed so that they can be refurbished multiple of times.