Scent marking entails significant energetic and opportunity costs that demand efficiency. Signal detection theory offers a theoretical framework that generates testable hypotheses regarding where animals should place scent signals in the environment in a way that maximizes their probability of detection by target receivers while minimizing costs of production and distribution. Solitary and reliant on chemical communication, the giant panda, Ailuropoda melanoleuca, offers an interesting model to test these ideas. We studied scent-marking patterns in wild giant pandas in the Foping Nature Reserve by surveying areas containing a high density of scent posts. Pandas did not deploy scent marks randomly in this environment, but targeted trees with specific characteristics that promoted signal persistence, range and/or likelihood of detection. Variables affecting selection of scent-marking sites included bark roughnesss, presence of moss on the tree trunk, tree diameter and distance to the trail. That pandas should be efficient with their use of chemosignals comes as no surprise, as mounting evidence is suggesting that many aspects of giant panda life history are constrained by their energetically poor diet. We also found seasonal and sex differences in marking patterns, indicating a role for scent marking in reproduction and competition. Males scent-marked throughout the year, whereas females scent-marked predominantly during the mating season, suggesting functional differences in scent marking between the sexes.