The extant laboratory research on anxiety-related attentional bias in response to external threat cues is discussed. Studies utilizing dichotic listening tasks, modified Stroop color naming tasks, and reaction time tasks are critically reviewed. Findings are summarized according to method, anxiety disorder, clinical versus nonclinical subjects, and treatment outcome studies. The data suggest that clinically anxious subjects show an attentional bias towards threat cues specific to their disorder, while anxious nonclinical subjects show bias towards general threat. However, methodological concerns preclude absolute statements. The implications of attentional bias are discussed, and recommendations for future research are made.