In this work, we develop a distributed-parameter model which is capable of predicting the dynamic response of bed height as a function of the salient properties of the solid and fluid phases. We show that a model which captures both the convective and dispersive nature of solid transport in a solid-liquid fluidized bed can adequately predict the bed-height evolution under both step increases and step decreases in the fluidization velocity. The convective term is described by an augmented form of the Richardson-Zaki expression. We propose a novel Bingham-like model to describe the diffusive characteristics of the solid-liquid system. It is shown that the dispersion scales as a function of the dimensionless column-average solid velocity, which is proportional to a yield stress. When this velocity is large, as in the initial period after a step change, the dispersion mechanism is not operable in the system. After the column-average solid velocity drops below a critical level, the dispersion mode is operable. A series of experiments were performed for both step increases and step decreases in the fluidization velocity. Model predictions for the bed-height transients are in good agreement with experimental data.