Separation-related problem behaviors (SRPB) is a common problem in dogs and a common cause of their relinquishment. Despite its severity and prevalence there is little research on the prediction of SRPB. Being able to predict SRPB in shelter dogs would help shelters identify at-risk dogs and allocate pre- and post-adoption resources to those dogs, including providing more targeted behavior counseling and better placement decisions. In this talk, we will discuss the current state of scientific literature on characteristics of dogs or owners that correlate with dogs exhibiting SRPB and then delve into our own behavioral study on predicting SRPB. We tested whether we could predict post-adoption behavior of stray shelter dogs from an in-shelter test. We tested 27 shelter dogs. After interacting with the dog for 30 min, we left it alone in the room and video-recorded its behavior. We coded behaviors associated with SRPB as well as those not associated with SRPB (e.g., play or passive behavior). Dogs additionally wore an activity monitor, and we collected salivary cortisol at three time points. Finally, we contacted adopters approximately 6 mos after adoption to determine dogs’ at-home behavior. We assessed the time-course of different behaviors of individual dogs across the 30 min test as well as the individual dogs’ time allocation between different behaviors. We will discuss the relationships between behavior on an in-shelter test, activity monitor data, changes in cortisol, and post-adoption behavior and what that means for predicting dogs’ needs in shelters and after adoption.