Vibration characteristics of agricultural machinery are generally complicated owing to so many signals vibrated by driving units, external forces, and so on. In this study, the wavelet analysis was performed in order to extract the specific vibration (or fluctuation) wave-form among complicated vibration (or flictuation) signals. First, the vibrated acceleration of cutting-section of the combine harvester and dynamic load fluctuation acting on the track rollers of the rubber crawler driving unit were measured. As a result, the former acceleration of cutting-section indicated complicated signals with a collision vibration of driving unit except for fundamental revolutionary vibration, and further, the latter dynamic load fluctuation indigated signal with periodical influence of rubber crawler joint except for dynamic load fluctuation. So, using the wavalet analysis, extraction of fundamental vibration and load fluctuation singals among all signals in the time domain was attemoted. As a result, specific vibration(fluvtuation) waveform could be clearly extracted by resolving into every frequency band through the disacceleration and rubber crawler's dynamic load were made clear by means of extracting from other noise signals. Further, waevlet analysis is convenient for analyzing complicated vibration characteristics involving so many noise signals.