In order to understand the characteristics of slug coalescence, the slug frequency has been measured by a differential pressure method in a gas-fluidized bed of i.d. 0.1 m and height 2.25 m, at ambient temperature and pressure, and with respect to variations of bed height (0.23-0.93 m) and gas velocity (U-U-mf = 0.0190-0.139 m/s). Air was used as fluidizing gas and a fresh FCC catalyst (Geldart's group A particle, d(p) = 0.070 mm, rho(s) = 1600 kg/m(3), U-mf = 0.0113 m/s) and glass bead (Geldart's group B particle, d(p) = 0.168 mm, rho(s) = 2486 kg/m(3), U-mf = 0.0303 m/s) as bed materials. The slope of a frequency decrease accompanying with a gas velocity increase at different bed heights was similar in the bed of Geldart's group-A particle. So was the slope of frequency decrease with bed height at different constant gas velocities. The mechanism controlling slug coalescence was found to be different between Geldart group-A and B particles. The slug coalescence model proposed by Yamazaki et al. (1984) was extended to the case with gas velocity as a variable. The slug coalescence rate constant leveled off after an initial decrease in the bed of Geldart's group-B particle. However, it increased in the bed of Geldart's group-A particle as the gas velocity increased. Correlation on the slug coalescence rate constant that includes effects of gas velocity, temperature and particle properties has been proposed in each different slug growth regime.