Mood effects on automatic and controlled components of semantic primin
g were examined. After a neutral or positive mood manipulation, subjec
ts underwent two blocks of a priming task. Automatic and strategic pri
ming processes were operationally differentiated by the assumption tha
t automatic priming remains constant during the task, whereas controll
ed priming develops slowly during the course of the experiment. In the
first block the priming effect was greater for happy than for neutral
mood participants. In the second block the priming effect was greater
for neutral mood than for happy participants. The results confirm the
hypothesis of a mood-dependent processing change: During early trials
priming is based on automatic processing facilitated by positive mood
. This processing mode is later superseded by controlled processes aid
ed by neutral mood.