AbstractRecent years have raised questions about the effectiveness of attentional bias modification (ABM) in individuals with social anxiety. In the current study, we employed a novel training method—ABM‐positive‐search training—to modify attentional bias in socially anxious individuals. The attentional bias was measured using the dot‐probe task, and brain electrical activity was recorded. The ABM‐positive‐search group was instructed to search for a happy face in a 3 × 3 visual array. The placebo control group was asked to find out whether the only male from the female or the only female from the male. In terms of behavioral indicators, the results showed a significant reduction in the variability of trial level‐bias score (TL‐BS) in the ABM‐positive‐search group post‐training. However, there was no significant change observed in the placebo control group. The electrophysiological results were consistent with the behavioral findings. To be specific, both groups displayed the N2pc effect in response to threatening faces before the training. After training, the N2pc effect disappeared in the ABM‐positive search group, while it remained in the placebo control group. In conclusion, ABM‐positive‐search training can effectively modify the attentional bias of socially anxious individuals, and cognitive control plays an important role in this process.