The present study sought to extend and refine the emotional priming pa
radigm described by Power and Brewin (1990). The paradigm involves the
presentation of emotion-related prime stimuli followed by negative or
positive self-descriptive adjective probes. The original study showed
that, following a negative prime, normal subjects were slower to endo
rse negative adjectives as self-descriptive at a long delay relative t
o a short delay. This was interpreted as evidence for some form of sel
f-esteem regulation process. The present study partially replicated ea
rlier findings with normal subjects. However, a clinically anxious gro
up (generalized anxiety disorder patients) showed a different pattern
of response. There was evidence of a priming effect in these individua
ls, ie, a processing bias in favor of negative stimuli The theoretical
implications of these findings are discussed.