Friction brakes represent the most important safety feature literally in all vehicles and their rigorous “friction testing” is usually performed on several platforms/scales and completed with field tests. Since friction and wear are system properties, it is not trivial to design “small scale” tests and to correlate data generated at different levels of testing complexity. Nevertheless, the economy of the brake materials development process could be improved, when interpretation of friction and wear test data is based on a deeper/proper understanding of physics and chemistry of ongoing friction phenomena. This contribution follows the two series previously presented at SAE Brake Colloquia and compares the data generated in the full-scale brake dynamometer SAE J 2522 performance test (Link Engineering 2800M dynamometer) with data generated in bench-top (small scale) friction tester (Bruker UMT) equipped with environmental chamber controlling temperature. Scaling laws of physics were adopted for design of the small-scale testing procedure, however, a different scaling philosophy as well as different friction materials were used when compared to the previously reported findings. Identical commercial OEM brake pad samples containing biodegradable environmentally friendly fibers and commercial OEM cast iron rotors were used in both dynamometer and scaled-down bench-top friction tests. Friction and wear surfaces/mechanisms were studied by using scanning electron microscopy (Quanta FEG 450 by FEI) equipped with the energy dispersive X-ray microanalysis (Inca System), and 3D optical microscope (NPFLEX by Bruker). Major conclusions proposed for this study can be summarized as follows: 1) Proper scaling by using physics principles allows for reasonable correlation of dynamometer and bench-top test data, although the results differentiate, particularly during fade and high temperature tests. These findings further support the previously published data and indicate that differences in scaling philosophy neither the types of tested materials have considerable impact on the generated data. 2) It is very important to properly select representative pad samples, as their size is considerably smaller compared to full pads. When the identical rotor materials are used, the repeatability of data is excellent and the sensitivity to typical differences of the bulk microstructure of cast iron is minimal. 3) When the testing results generated on dynamometer and bench tester matched well. the friction surfaces of full pads tested in dynamometer and the friction surfaces of small pad samples exhibited identical topography and chemistry.