This is a preliminary analysis of digitized satellite cloud brightness data that will be used in developing a model for large-scale analysis over data-void tropical areas, which is based on Wallace's (1971) postulation that such data may be used to estimate vertical motion in the tropics. The relationship between satellite cloud brightness and large-scale 200mb divergence on 5° x 5° grids in the tropical western North Pacific is examined from several angles. Correlation coefficients between the two fields are computed on a daily basis for the region of study and at each grid point for 8-month time series. The coefficients are mostly low although a positive correlation is generally indicated. The low correlation may be due to the quality of the available data, especially for the divergence which is kinematically computed from analyzed winds. Spectrum analysis is also performed to determine dominant synoptic time scales for both parameters in the period of study. Two common period bands, one centered ~ 10 days and the other ~ 5 days, are found. Cross-spectra between the two parameters indicate that their phase differences are small (< 1/4 cycle) for both bands, but the coherence squares are somewhat lower than those found previously by Wallace for the 4-to 5-day waves using direct radiosonde data. Horizontal structures of both parameters, determined from inter-longitude cross-spectra in the two bands, suggest that the brightness is much more organized than divergence, but the two fields resemble each other whenever an organized pattern in divergence can be found. The results of this study lend some support to Wallace's postulation, but also indicate some problems for using the present data. It is anticipated that the correspondence with the actual divergence field will be greatly improved if satellite infrared data is incorporated into the brightness data. (Author)