Background and Objectives. Across three experiments we investigated transfer effects of single-session attention bias modification via dot-probe training. Methods. In experiment 1, participants received training either toward or away from negative images or no-training, and transfer to an affective task-switching task was examined. In two other experiments, participants were trained to orient attention toward either positive or negative words (experiment 2a) or facial expressions (experiment 2b), and transfer to an interpretation bias task was examined. Results.In all experiments, the dot-probe training procedure did not effectively modify biases in attention allocation at the training condition level, but produced a large variability in individual attention bias acquisition within and across conditions. Individual differences in pre-training attention bias and attention bias acquisition were not related to performance on the affective task-switching task or the interpretation tasks. Limitations. The present investigations are limited by the lack of effectiveness of ABM at the condition level, the order in which transfer tasks were administered, and the restricted range of affective symptoms that could moderate training and transfer effects.