This paper presents a vibration based fuzzy classifier used to detect fault conditions in bearings. The fuzzy knowledge foundation is obtained from the behavior of the vibration signal in presence of a fault condition in outer race, inner race, cage and rolling element. The fuzzy system is implemented in Matlab® and it is validated in a vibration dedicate test bench used for research proposes. The system performance exhibits reliability and robustness in the fault detection in each experimental fault scenario.