Purpose: Our purpose was to estimate a threshold value for change for the six dimensions of the Impairment Inventory of the Chedoke-McMaster Stroke Assessment and the confidence in labelling a person as having improved or not. Method: Secondary analysis of two data sets, previously reported by two research teams, consisted of two statistical analyses. The first analysis used a multiple of the standard error of measurement to calculate the threshold value for change for the six dimensions. The second analysis used the diagnostic test method to calculate a threshold improvement value and the confidence a clinician had in labelling a person as having improved or not on the leg, foot, and postural control dimensions. Results: The threshold value for change was determined to be 1 impairment point (i.e., stage) for the arm, hand, leg, foot, and postural control dimensions and 2 impairment points for the shoulder pain dimension. The positive predictive values associated with the leg, foot, and postural control dimensions were 74%, 59%, and 65%, respectively. Conclusions: Clinicians can use a change of 1 impairment point for the arm, hand, leg, foot, and postural control dimensions and a change of 2 impairment points for the shoulder pain dimension to identify true change in a patient's motor recovery.