For the Timed 25-Foot Walk (T25FW) and 9-Hole Peg Test (9HPT), components of the Multiple Sclerosis Functional Composite (MSFC), cut-off points of 20% change have previously been defined as meaningful endpoints of functional decline. Recently, however, a 15% change of MSFC components was introduced. The objective of this study was to determine optimal cut-offs for all MSFC components to indicate clinical disease progression in a primary progressive (PP) multiple sclerosis (MS) population. T25FW, 9HPT and the Paced Auditory Serial Addition Test (PASAT) were performed in 161 patients with PPMS with a 2-year interval. Absolute and relative differences in test scores were calculated. For each cut-off point of relative change, proportions of patients who progressed (deterioration beyond cut-off value) and improved (improvement beyond cut-off value) were calculated. Further, we calculated the ratio of 'improved' versus 'progressed' patients. Line graphs were created indicating: percentage progressed patients, percentage improved patients, and ratio of improved versus progressed patients. The optimal cut-off was determined by searching the cut-off point with the lowest ratio of improved versus progressed patients, while at the same time capturing a substantial amount of progression. For both T25FW and 9HPT, the ratio between patients that improved and worsened clearly decreased between the cut-offs of 15% and 20%. For the PASAT, the ratio between patients improved and worsened was persistently poor. In conclusion, a cut-off of 20% for both T25FW and 9HPT has a better signal-to-noise ratio than lower values (e.g. 15%) and is therefore preferable for the assessment of disease progression. No satisfactory cut-off point for the PASAT could be determined.