DescriptionAttention bias, or increased attention to smoking relative to neutral cues, may be a useful marker for relapse risk; however, attention bias has not been consistently documented in continuing smokers which limits its utility for treatment planning. This study assessed attention bias in continuing smokers experiencing a mild stressor. Attention narrowing, or increased attention to smoking relative to other positive cues, was also studied to assess the specificity of distress effects on attention to smoking cues. Analyses also explored the degree to which attention bias and narrowing predicted smoking lapses. In addition, because the mild stressor used was somewhat novel, we tested the degree to which unpredictable noise induced attention bias toward negative cues and subjective distress relative to control. Data were collected in two similar study designs in smokers preparing for a quit attempt. Designs differed in the control conditions against which stress was compared within subjects (Design 1 = predictable noise, Design 2 = silence). Attention was assessed using a modified Stroop task with neutral, negative, positive, and smoking words. Design, Stress Order (whether unpredictable noise was presented first or last), and Word Order (whether smoking words were presented first or last) were included in analyses to explore their potential effects on attention. Results indicated that attention bias effects were greater in Design 1 than Design 2 and when stress was presented first rather than last, although attention bias toward smoking cues was small in magnitude in all conditions. No evidence for attention narrowing was found. Attention bias and narrowing under mild stress were not related to later smoking lapses. Although the stress manipulation did not increase distress relative to control, it corresponded with a negative bias as anticipated. The magnitude of this effect varied across study designs. Overall findings suggest attention bias effects are not robust among continuing smokers experiencing unpredictable noise stress, and effect sizes vary as a function of design elements in important ways. Further research is necessary to determine the optimal context to reliably assess attention bias in continuing smokers.