Abstract Objective: Two scoring systems for a verbal fluency test were compared using the Rasch Rating Scale Model. Method: The analysis was carried out on 289 participants, 92 of whom had had a Parkinson’s disease diagnosis. Scores were calculated with two different category systems: a conventional procedure and a percentile-based one. Results: The percentile-based Rasch scores produce adequate categories and reliable measures, while the correlation with the Mini Mental State Examination evinces concurrent validity. After statistically controlling for age, percentile-based Rasch measures discriminated between both groups, demonstrating predictive validity. Conclusions: The analysis of the two procedures allows for the recommendation of the use of percentile-based categories.