Abstract Ninety 6- and 7-year-olds (49.3% White, mostly middle class) from greater Washington, DC were randomly assigned to a subliminal priming condition (secure, happy, or neutral) to determine if attachment security priming decreases physiological, expressive, and self-reported fear reactions to threatening stimuli. Dispositional attachment security was also assessed. Secure priming and attachment security each decreased electrodermal reactivity, increased vagal augmentation, and decreased fearful facial expressions compared to control conditions. Examination of a statistical interaction between security priming and child attachment indicated that, although secure children had increased vagal augmentation and fewer fearful expressions than insecure children, the effects of priming were constant across secure and insecure children. There were no priming or attachment effects associated with children's self-reported fear.