This paper presents results of experimental investigation into the creep groan problem of disc brakes when used on rear axles of cars and light trucks. Improvements in test procedures and data analysis leads to clearly defined limit cycles in the motion of brake components. The limit cycles are related to both brake design and axle resonances. Creep groan frequency and signature are shown to depend on the choice of friction material. Simple monotonic friction models are shown to be limited in their capability of describing creep groan motion.