Attachment styles are often primed by having participants recall and describe a relationship that is prototypical of a given attachment style. Researchers may exclude participants who cannot recall such a relationship, or who describe relationships that do not conform to the assigned prime. I suggest that excluding participants is untenable, and may threaten a study’s validity. In the present research, I examine predictors of exclusion from an attachment priming study. Priming insecure attachment resulted in greater odds of exclusion relative to a control condition. Female participants with greater sexual experience also had lesser odds of exclusion. These results suggest that attachment-priming procedures contribute to participant exclusion that compromise internal and external validity. Discussion focuses on directions for future attachment-priming research.