This study investigated whether individuals with a high trait anxiety (HTA) rating retain attentional bias towards threatening pictures when such pictures are repeatedly presented. Subjects rated high (n = 20) and low (n = 20) on an anxiety scale participated in a forced-choice reaction time version of a modified dotprobe task. Picture pairs were presented in four exposure blocks. On each exposure, the attentional bias to the threatening pictures was measured. HTA individuals showed more selective attention to the high threatening pictures than individuals with low trait anxiety (LTA). However, HTA individuals did not maintain attentional bias. On the fourth block, the attentional bias to threatening pictures disappeared. Theoretical and methodological implications of these findings are discussed.