The most robust demonstrations of the nicotine-related performance effects on human cognitive processes are seen in tasks that measure attention. If nicotine does have some potential for enhancing attention, the obvious question to ask is whether the effects demonstrated in the laboratory hold any significance for real-life performance. This paper describes three studies that compare the effects in smokers of a single own brand cigarette on laboratory tests of attention and on everyday analogues of these laboratory tasks. In the laboratory measures of sustained attention and in the everyday analogue, performance advantages were registered in the smoking condition. These benefits were observed in smokers who abstained for a self-determined period of not less than 2 h. The studies were unable to replicate previous research reporting positive effects of smoking on a laboratory task of selective attention, the Stroop task. Small but significant improvements in performance were registered in the everyday analogues, which involved sustaining attention in a dual task situation, a telephone directory search task and a map search task. In addition, smokers showed a significant colour-naming decrement for smoking-related stimuli in the Stroop task. This attentional bias towards smoking-related words occurred independent of whether they had abstained or recently smoked an own brand cigarette. The effect is discussed in terms of the two-component model of processing bias for emotionally valenced stimuli.