Aim: To assess the responsiveness, concurrent validity, and feasibility of the Dyskinesia Impairment Scale (DIS) in non-ambulatory patients with dyskinetic cerebral palsy (CP). Method: The study is a secondary analysis of data collected in the IDYS trial, a randomized controlled trial on the effects of intrathecal baclofen (ITB). The DIS and Barry–Albright Dystonia Scale (BADS) were conducted at baseline and after 3 months of ITB or placebo treatment. Responsiveness was assessed by comparing the effect sizes and correlation of change after treatment between the DIS and BADS. Concurrent validity was evaluated by assessing the correlations between scales. Feasibility was evaluated for each DIS item by the number of participants who successfully accomplished the item. Results: Thirty-three non-ambulatory patients (9 females, 24 males) with dyskinetic CP (ITB-treated: n=17, mean [SD] age: 14y 1mo [4y 1mo]; placebo-treated: n=16, mean [SD] age: 14y 7mo [4y]) were included in the study. The effect sizes for BADS and DIS were similar in The ITB-treated group (−0.29 and −0.22 respectively). Changes after treatment on the DIS dystonia subscale correlated with changes on the BADS (r=0.64; p<0.001). The DIS dystonia subscale and BADS correlated at baseline and follow-up (r=0.78; p<0.001 and r=0.79; p<0.001). Not all DIS activity items could be performed in this sample of patients. Interpretation: For non-ambulatory patients with dyskinetic CP, the responsiveness of the DIS equalled the responsiveness of BADS. Concurrent validity was adequate. Feasibility for activity items was restricted in patients with severe dyskinetic CP.