Two-hundred and nine normal subjects underwent a study aimed at providing norms for some attention tests currently used in neuropsychological examination. Norms were calculated taking into account the demographic variables. Our battery included single visuo-manual and go/no-go reaction times, and two versions of the Stroop Colour-Word Test. For the latter test, we employed a traditional version based on card presentation and a new version based on computer-assisted presentation. We evaluated the intercorrelation of the different tests, and studied the statistical regression models in order to adjust the original scores according to age, education and gender. Normative values were calculated following the Equivalent Scores method. The pattern of intercorrelation of our battery and the regression models of the scores are discussed with reference to the literature. The newly developed computer-assisted presentation of the Stroop test offers some advantages over the traditional version, especially with patients affected by spatial disorders.