Background/Aims: Cognitive dysfunction is a major handicap in multiple sclerosis (MS). Its prevalence varies due to disease heterogeneity and methodological issues. A neuropsychological battery of intermediate size was designed for and explored in the screening of cognitive dysfunction in MS patients. Methods: The battery was administered to a hospital-based sample of 191 MS patients and 50 matched controls. Eleven test scores measuring verbal fluency, verbal learning, attention, calculation and visuoperceptual ability were selected on the basis of sensitivity and lack of redundancy. Two alternative approaches were compared for diagnosis of cognitive dysfunction based, firstly, on the number of failed tasks, and secondly, on a single standardized global score. Results: The approach based on the number of failed tasks discriminated better than did the global approach between patients and controls. Using a cutoff of two altered scores, a cognitive dysfunction prevalence of 34% was obtained. The score yielded after summing errors in all tests was the most frequently altered and proved particularly useful for detecting minimally impaired patients. Conclusion: The purpose-designed battery was adequate for the screening of cognitive dysfunction in MS patients. The better accuracy of the single-task approach might reflect MS heterogeneity.