Primary objective: In a child with pure ataxia after tumour removal, does dexterity training at the right elbow improve dexterity at the right elbow? Does improvement in dexterity transfer to improvement in upper limb activity? Research design: Single-case experimental study with 2-week baseline, 2-week intervention and 2-week follow-up phases and assessment by a blinded assessor. Methods and procedures: A 5-year old child 3 years after surgical resection of a low-grade cerebellar tumour underwent to dexterity training. Dexterity was measured using a Finger-to-Nose Test and upper limb activity was measured using the 9-Hole Peg Test. Measures were collected every 2 or 3 days over the 6-week period of the study. Results: Finger-to-Nose Test showed visible but not statistical improvement between baseline and intervention phases in terms of smoothness of movement. On withdrawal of the intervention, visual analysis showed that scores remained at the same level. The 9-Hole Peg Test showed visible but not statistical improvement in terms of speed and smoothness of movement. On withdrawal of the intervention, visual analysis showed that improvement in speed was maintained during follow-up, but smoothness decreased. Conclusions: There is some evidence that dexterity can be trained and that improvement carries over to activity.