This project was designed to understand the causes and mechanisms of bicycle disc brake noise and use that information to formulate and evaluate possible mitigation techniques. Brake noise was generated by a real bicycle running on a treadmill and recorded by microphone and laser vibrometer. Six independent variables, brake force, rotor thickness, front fork stiffness, weather conditions, spoke tension, and friction coefficient, were varied according to a one-quarter fractional factorial design. A finite element model of the rotor, pads, and calliper was also formulated and analysed. The results of these two methods, particularly the disc mode shapes and frequencies, suggest that doublet mode splitting and reconverging plays a role in noise generation and that changing the rotor mass or breaking its symmetry could interfere with such noise generation. Finally, of these mitigations, breaking disc symmetry proved the most fruitful, with noise magnitude reductions from 72% to 99%, depending on frequency.