Wall vibration of a lab-scale fluidized bed was measured at various gas velocities (0.4-1.4 m/s) and aspect ratios (L/D: 0.5, 1.0 and 1.5), filled with 1500-mu m particles, as well as that of the empty bed. For comparison with the vibration signals, pressure fluctuations inside the bed were simultaneously collected. It was shown that the vibration intensity of the empty bed is significant compared to that of the fluidized bed. Thus, the original vibration signal was decomposed into coherent signal (reflecting the vibration arises from the empty bed) and inherent signal (reflecting the vibration arises from the bed hydrodynamics). The analysis of inherent signal showed that this measuring technique in conjunction with signal components decomposition are robust enough for detecting sharp changes (such as de-fluidization) in the bed hydrodynamics. The same results were obtained from pressure fluctuations. The vibration signal analysis showed that the de-fluidization velocity falls within the range of 0.44 and 0.47 m/s and this value is within the range of 0.44 and 0.46 m/s for pressure fluctuations. The frequency range of different length-scale phenomena was determined. The frequency ranges of macro-, meso-, and micro-structures were, respectively, 0-5, 5-40, and 40-200 Hz for pressure fluctuations. These frequency ranges were 0-9500, 9500-12 000, and 12 000-32 500 Hz for vibration signal. The vibration of bed wall qualitatively reflects the gradual changes in bed hydrodynamics, like changes in the relative energy of bubbles and clusters. However, it was shown that pressure fluctuations are more sensitive to these gradual changes.