Background <p>The Stroop Colour Word Test (SCWT) is a neuropsychological tool widely used to assess the ability to inhibit cognitive interference, particularly in patients with dementia. This study aimed to provide normative data for a short version of the SCWT in an Italian sample of 452 healthy individuals aged 20–90 years (education ≥ 5 years).</p> Methods <p>The test version included 50 items using four colours (yellow, red, green, blue) and involved four tasks: (1) reading colour names in black ink, (2) naming coloured dots, (3) reading colour names, and (4) naming the ink colour of incongruent colour words. For each task, response time and errors were recorded. To analyze performance, we developed linear models including all combinations of the independent variables (sex, age, education), applying transformations (square root, logarithm, reciprocal, quadratic, cubic) to account for non-linear effects.</p> Results <p>In total, 98 models per score were generated and assessed using the Bayesian Information Criterion (BIC). Correction factors were derived by comparing predicted scores from the best-fitting models to observed averages. The most accurate predictive models were identified for each performance index.</p> Conclusions <p>The availability of these normative values enhances the clinical utility of the short SCWT in evaluating selective attention and interference control in both adults and older adults. This updated dataset contributes to refining neuropsychological assessments in this cognitive domain.</p>