The present research explored the effects of Situational Attribution Training (Stewart, Latu, Kawakami, & Myers, 2010) on affective bias utilizing facial electromyography (EMG). Participants viewed a slideshow of randomly presented photographs of both and White and Black American men while rating how “friendly” each individual appeared. Simultaneously, corrugator and zygomaticus region activity, linked with positive and negative affect, respectively, was measured. Of these participants, half were randomly assigned to complete Situational Attribution Training beforehand. Results for EMG activity suggested no significant differences in EMG activity for White compared to Black photographs for either the training or control participants; thus, this study did not find evidence of affective bias by way of corrugator or zygomaticus activity. However, errors in slideshow presentation prevent clear interpretation of these results. Suggestions for future research and ways in which bias errors can be avoided are discussed.