The relationship between pressure distribution and cavitation (noise) inside throttling groove is investigated by numerical simulation and experimental method. A valve pocket with several transducers is performed to detect the pressure distributions inside the valve chamber, and the results fit quite well with the computational fluid dynamics(CFD) analysis. High-speed imaging techniques are employed to investigate the cavitation mechanisms, in particular bubble inception and cluster formation near the throttling groove. A spectrum analyzer is used to measure the sound pressure level of noise generated by the bubble flow. It is found that the pressure distributions inside the groove are sensitive to the valve port configuration and back pressure. The pressure distribution determines the bubble size and number passing through the valve grooves and the sound pressure level of noise induced by collapsing bubbles. The inlet pressure mainly affects the saturation degree of bubbly flow inside the groove and the intensity of sound pressure level accordingly.