[eng] The main objective of this study was to compare children select-ed in a community setting with ADHD, high-performance, and normo-typical development on a sustained attention task. Three groups of chil-dren were selected: ADHD (n = 42), high-performance (n = 20), and nor-mo- typical development (n = 28). A brief computerized vigilance task (CSAT-R) was applied to compare attentional capacity and reaction time. The participants were divided into those with “attentional dysfunction” and those with “normal attention” to analyze clinical validity. Children with high-performance were clearly differentiated from the other two groups, with large effect sizes. The differences between normo-typical and ADHD groups were only significant in the errors and in a nonparametric index of attentional capacity, but with small effect sizes. The CSAT-R showed good specificity and an acceptable positive predictive value, but low levels of sensitivity, and a poor negative predictive value. Therefore, sustained attention could be a prominent mechanism in children with high capacities. The CSAT-R (and probably most of attentional tasks) would be moderately useful in community settings for ADHD diagnosis, but not to rule it out.