As floodplain soils are often contaminated, we studied the release of trace metals from three topsoil horizons in column experiments with variable flow interruptions and flow velocities, compared it with that in batch leaching tests and evaluated the column data by inverse simulations. Only small proportions (<1%) of trace metals present in the neutral and humic soils were mobilised by the batch leaching tests and the column experiments. Release of Cr, Cu, Ni and Zn in the column experiments was rate-limited, as detected by increased concentrations after flow interruptions. A combination of linear equilibrium and non-equilibrium isotherms reflected the Ni and Zn elution data, with Zn release being slower. Simulated values for initially bound metals available for release are in the same order of magnitude as those determined by the batch leaching tests. However, the consistency of both experimental approaches decreases with increasing rate limitation, as detected here for Zn.