The paper illustrates the high degree of correlation now attainable with practical vehicle test results, using a reduced-scale dynamometer. The techniques for achieving this state-of-the-art are described in relation to the parameters generally known to influence vehicle/laboratory correlation. As a result of the investigation, the following additional parameters are submitted: history of pad use, correct temperature/time simulation of the car by the machine, and rate of work input during a test. The simulation item is the subject of a special study in which the thermal conditions in a brake are examined on an analog computer by means of a 2 d.f. model.A discussion of the possible applications of the scale dynamometer leads to the conclusion that a machine could now be developed to provide meaningful data on linings for the use of friction material, brake, and vehicle manufacturers, as well as legislators.