Background: Determining the level of manual ability of children with cerebral palsy has significant role in scheduling care and providing supportive services by organizations such as the a social welfare office. Manual Ability Classification System (MACS) is responsible for this critical matter. In Mazandaran Province, the prediction of manual ability is intuitive and is done without tools. This study aimed to investigate the consistency between operational therapists prediction of future manual function of children with cerebral palsy and MACS scale.

Materials and Methods: This study was designed as a cross sectional trial. The study population consisted of 12 occupational therapists working in Mazandaran’s rehabilitation center under the social welfare office. Firstly, occupational therapists chose 100 children with cerebral palsy and classified their future manual ability into five levels according to their prediction. Then the researcher classified the children based on MACS scale. Finally, the amount of prediction consistency between therapists and researcher was statisticaly analyzed.

Results: The weighted kappa coefficients of MACS scale were 0.671 in first level and 0.747 in fifth level that show good agreement in these two levels. This coefficient was 0.417 in third level and 0.444 in forth level that shows fair agreement. The weighted kappa coefficient was 0.358 in second level that indicates slight agreement.

Conclusion: With prediction consistency between operational therapists and researcher, MACS is used as a suitable means for classifying the level of manual ability in children with cerebral palsy and predicting their needs to adaptive and auxiliary devices by occupational therapists in rehabilitation centers and social welfare offices.