There is a growing interest in the potential benefits of attentional bias modification (ABM) training in chronic pain patients. However, studies examining the effectiveness of ABM programs in fibromyalgia patients have demonstrated inconclusive effects on both behavioral indices and clinical symptoms. Additionally, underlying neural dynamics of ABM effects could yield new insights but remain yet unexplored. Current study, therefore, aims to investigate the effects of ABM training on known neural electrophysiological indicators of attentional bias to pain (P2, N2a). Thirty-two fibromyalgia patients were enrolled and randomly assigned to an ABM training (N = 16) or control (N = 16) condition (2 weeks duration). Within the ABM training condition participants performed five sessions consisting of a modified version of the dot-probe task in which patients were trained to avoid facial pain expressions, whereas in the control group participants performed five sessions consisting of a standard version of the dot-probe task. Potential ABM training effects were evaluated by comparing a single pre- and post-treatment session, in which event-related potentials (ERPs) were recorded in response to both facial expressions and target stimuli. Furthermore, patients filled out a series of self-report questionnaires assessing anxiety, depression, pain-related worrying, fear of pain, fatigue and pain status. After training, results indicated an overall reduction of the amplitude of the P2 component followed by an enhancement of N2a amplitude for the ABM condition compared to control condition. In addition, scores on anxiety and depression decreased in patients assigned to the training condition. However, we found no effects derived from the training on pain-related and fatigue status. Present study offers new insights related to the possible neural mechanisms underlying the effect of ABM training in fibromyalgia. Clinical trial (TRN: NCT05905159) retrospectively registered (30/05/2023).