BACKGROUND AND OBJECTIVES: Attentional Bias Modification (ABM) may constitute a new type of treatment for affective disorders. ABM refers to computerized training programs that have been developed based on laboratory findings in experimental psychology. Meta-analyses have reported moderate effect sizes in anxiety disorders. Two small studies have also claimed an effect in dysphoria. Furthermore, a series of studies in individuals with low self-esteem has shown that they benefit from a single session of an ABM variant based on a visual search task. The current study tested the working mechanism of visual search ABM in dysphoria. METHODS: Forty dysphoric individuals engaged in a single session of ABM training or control training. Attentional bias for positive and negative facial expressions was assessed pre- and post training. Positive and negative mood states were assessed throughout the procedure. RESULTS: Attentional training had no effect on attentional bias. Positive and negative mood states were not differentially affected by training condition. LIMITATIONS: Small treatment effects may have gone undetected and there are some methodological differences with prior research. CONCLUSION: We found no evidence that engaging in a single session of a visual search ABM modifies attentional biases for happy, sad or disgusted facial expressions.