It was hypothesized that smoking's calming effects are cognitively med
iated and depend on the presence of a benign distractor; smoking narro
ws the focus of attention, thereby reducing anxiety by facilitating di
straction from stressful cognitions (cf. C. M. Steele & R. A. Josephs'
s [1988] attention-allocation model of alcohol reinforcement). This no
tion was tested by examining the effect of smoking (vs. not smoking) o
n anxiety with and without a concurrent distraction in 82 smokers; dis
traction effects were also assessed in 42 nonsmokers. As predicted, sm
oking reduced anxiety only when paired with a distracter. Further, the
se findings could not be explained by direct nicotine effects or nicot
ine withdrawal. Several measures of attention allocation failed to con
firm the hypothesized cognitive mechanisms, however. Implications of t
he findings are discussed.