Impaired response inhibition plays a major role in many addictive behaviors. However, in studies using go/no-go tasks, findings regarding the presence of response inhibition deficits in nicotine-dependent individuals are mixed. This might be due to differences between studies on a number of task parameters. Here we aimed to identify task conditions under which go/no-go task performance deficits can be observed in smokers and to characterize the nature of such deficits. Sixty-one male students (30 smokers, 31 non-smokers) performed a go/no-go task while independently manipulating three task parameters: (1) percentage no-go trials (50% or 25%), (2) stimulus presentation time (600 ms or 200 ms), and (3) nature of no-go stimuli (cigarette related or cigarette unrelated). Three measures, reaction time on go trials and percentage correct responses on go and no-go trials, served as performance indicators. Under 200-ms but not 600-ms stimulus presentation conditions, the smokers responded faster on go trials and made more errors on both go and no-go trials than the non-smokers did. These differences occurred irrespective of the percentage of no-go trials and nature of no-go stimuli. The accuracy differences disappeared after controlling for the response time differences, suggesting a strong speed-accuracy trade-off. This study contributes to unraveling the conditions under which smokers display impaired inhibition performance and helps to characterize the nature of this impairment. Under task conditions prompting fast responding, smokers are more prone to increase response speed and to make more errors than non-smokers.