Liquid fluidized-bed reactors are being investigated for the biological conversion of coal to clean-burning liquid fuels. The efficient design, operation, and scaleup of such reactors require accurate, predictive mathematical models which describe their performance. A fully predictive model is proposed to describe both the bed height and the particle size distribution obtained in such liquid fluidized beds of coal particles. This model is validated on both the macroscopic and microscopic scales. First, using well-characterized charges of Illinois #6 bituminous coal having a particle size of 10-250 mu m, the bed was studied macroscopically by observing the bed height as a function of Flow rate and coal charge. Secondly, to characterize the bed on a microscopic level, a novel, noninvasive fluorescence technique was used. This technique allowed direct observation of particle size distributions as a function of axial position and flow rate. Use of this experimental technique revealed patterns in bed segregation and elutriation that will be paramount to the operation of a bioreactor for coal solubilization. Over a wide range of particle sizes, liquid velocities, and coal loadings, the predicted bed heights agree very well with the experimental results. In addition, agreement is demonstrated between the predicted and measured particle number distributions as a function of flow rate and axial position. The mathematical model proposed here will have an impact on many low-Reynolds-number fluidized-bed applications, including other bioprocesses such as ethanol production and fermentation of organic acids.