In three experiments, we tested the hypothesis that negative priming (NP) c
an occur without prime or target selection, when conflicting properties are
associated with the prime and the target, and when the experimental condit
ions allow the encoding of the target as a separate episode from the prime.
These predictions were confirmed in Experiment 1, using a gender decision
task. Responses were slower when prime and target had the same gender than
when they had different genders, with an inter-stimulus interval (ISI) of 6
00 ms but not with an ISI of 25 ms. Experiment 2 eliminated a possible expl
anation of the NP obtained in Experiment 1, in terms of response inhibition
during the prime processing. Finally, Experiment 3 demonstrated the replic
ability and the generality of our NP, in a semantic categorization task. Em
pirical and theoretical consequences of our results for studies using the p
riming paradigm are discussed. (C) 2001 Elsevier Science B.V. All rights re
served.