One of the major challenges in disability management research is to express results in a manner that can be generalized to subjects with varying degrees of disability. Absolute measurements of change are often dependent on initial characteristics, in which case they can only be generalized to subjects with the same characteristics. We define proportional change as the ratio of change to the maximum possible or targeted change. It can be assessed in any situation where a maximum possible or targeted change is definable. Its estimated value will be sensitive to the choice of denominator. Subjective assessments, such as those measured with Likert scales, are naturally expressed as proportional change with the denominator being set by the subject. Denominators may also be determined objectively by physical limitations or, less desirably, by the measurement tool. Where there is no readily or objectively determinable denominator, they should be chosen for each subject in advance of the intervention according to carefully specified criteria. Proportional change is proposed, as an adjunct to the reporting of absolute measures of change following therapeutic interventions, as a means of expressing change in a manner that is both individualized and generalizable.