The purpose of inter-animal communication is to allow signals released by the animal to be perceived by others. Scent marking, with its characteristics of specificity and delay, is thought to be the primary mode of communication in giant pandas (Ailuropoda melanoleuca). The “signal detection theory” predicts that animals choose the substrate and location of their scent marks so that the signals released are transmitted more widely and remain longer. As an energetically marginal species, the cost of generating and marking chemical signals is costly for pandas, so they are predicted to make trade-offs in scent marking. However, existing studies do not account for the selective preferences of pandas for marking, as they are only explained by the density of marks at a certain location. Our study wanted to investigate whether the marking behavior of pandas is indeed consistent with signal detection theory. For the first time, we propose to use fecal counts to reflect the intensity of habitat use by pandas, combined with mark counts to determine the selective preference for marking. Our findings show that the scent marking behavior of pandas is consistent with signal detection theory, and that they go out of their way to ensure that their marks are detected. The results of the study will help us to further develop the conservation of pandas and their habitats.