International audienceWe consider a remanufacturing system with two streams of returned products and different variability levels (high and low). The arrival of returns with high variability is modeled with a hyperexponential renewal process and that of returns with low variability is modeled with a Poisson process. The reman-ufacturing facility can process the returned products in two ways. For the first way, each type of returns is remanufactured by a dedicated capacity. For the second way, returns from two different markets are remanufactured by a merged capacity. Analytical queueing models with the time value of money consideration are proposed for the admission decision, which decides on the acceptance or not of returned products based on quality and processing time. The proposed modeling determines the admission decision threshold value in order to maximize the total expected profit of the remanufacturing system. Our analysis also allows to study the interaction between the overall utilization and the arrival process variability. The results show the impact of the model parameters on the admission decision value and the total expected discounted profit. Also, the total expected discounted profit under the separated and merged capacities are compared