Purpose. The study aimed to establish whether the manual ability classification system (MACS), a valid classification system for manual ability in children with cerebral palsy (CP), is applicable in young adults with CP and normal intelligence. Subjects. The participants (n = 83) were young adults with CP and normal intelligence and had a mean age of 19.9 years. Method. In this study, inter observer reliability of the MACS was determined. We investigated relationships between the MACS level and patient characteristics (such as the gross motor function classification system (GMFCS) level, limb distribution of the spastic paresis and educational level) and with functional activities of the upper extremity (assessed with the Melbourne assessment, the Abilhand questionnaire and the domain self-care of the functional independence measure (FIM)). Furthermore, with a linear regression analysis it was determined whether the MACS is a significant determinant of activity limitations and participation restrictions. Results. The reliability was good (intraclass correlation coefficient 0.83). The Spearman correlation coefficients with GMFCS level, limb distribution of the spastic paresis and educational level were 0.53, 0.46, and 0.26, respectively. MACS level correlated moderately with outcome measures of functional activities (correlations ranging from -0.38 to -0.55). MACS level is, in addition to the GMFCS level, an important determinant for limitations in activities and restrictions in participation. Conclusion. We conclude that the MACS is a feasible method to classify manual ability in young adults with CP and normal intelligence with a good manual ability.