Objectives. The purpose of this study was to investigate whether cogni
tive bias for threat in anxious individuals mag be attributed to a def
ective inhibition. Design. It was investigated whether phobics and not
non-phobics would show a defective inhibition of threat words as comp
ared to neutral words. Method. Inhibition was measured by a negative p
riming task, which was administered to spider phobic participants (N =
29) and non-phobic controls (N = 31). Results. The phobics did not sh
ow less negative priming of threatening information. Instead, they sho
wed a general delay on all probe displays that were presented after th
reatening primes. Conclusion. Anxiety is related to defective inhibiti
on of threat, in the sense that this inhibition consumes extra cogniti
ve resources.