Study design: Longitudinal cohort study. Objectives: Previously, the Van Lieshout hand function test for tetraplegia (short form)(VLT-SF) has been transformed into an interval scale (r_VLT-SF) using the Rasch analysis, thereby fulfilling the requirements of an objective measurement. The present study aims (1) to transform r_VLT-SF logit scores (r_VLT-SF logit ) into 0-100 scores (r_VLT-SF 100 ) in order to facilitate communication amongst its users and (2) to describe r_VLT-SF 100 reference values in order to refine the prediction of upper extremity skilled performance in persons with tetraplegia. Setting: Eight Dutch rehabilitation centres. Methods: The VLT-SF data of tetraplegic patients from eight Dutch rehabilitation centres, gathered at the start of the active rehabilitation (t1), 3 months after t1 (t2) and at the time of discharge (t3), have been used. The r_VLT-SF 100 total score for t1 and t3 were computed, anchored on the data for t2. Reference values (medians and interquartiles) for different subgroups were calculated at different moments during the rehabilitation. Results: Data at t1 (n ¼ 64), t2 (n ¼ 73) and t3 (n ¼ 69) were used. The r_VLT-SF logit scores ( À4.78-6.32) were transformed into r_VLT-SF 100 (0-100) at t2. Reference values were established for persons with motor complete or incomplete lesions with either a high (C1-C6) or a low (C7-T1) lesion level at t1, t2 and t3. Significant differences (Po0.025) in r_VLT-SF 100 were found for most subgroups and different time intervals. Conclusions: The r_VLT-SF logit scores were converted into r_VLT-SF 100 scores, and reference values of the r_VLT-SF 100 have been established for different subgroups of tetraplegic patients at different stages during rehabilitation. The r_VLT-SF 100 has been proven to be able to detect improvement over time.