Three process-oriented dynamic acidification models were applied to a long-term monitoring site without calibration to evaluate the influence of model structural differences on simulation. The models were simplified to share as many commonalities as possible so that the main structural differences could be investigated. The models differed in sub-models for cation exchange, organic acids and acid anion speciation. All models were populated with ‘equivalent’ parameters by systematic input mapping. The influence of input variability was addressed through Monte Carlo parameter sampling. The three models behaved exactly the same for tracers (e.g. sulphate and chloride), indicating successful cross-parameterization of the models. Differences in model structure had an impact on some of the simulated chemical parameters. In particular, models using Gapon cation exchange simulated higher base saturation levels in the long run than their Gaines-Thomas counterparts, but simulated lower base cation concentration and acid neutralizing capacity in soil solution when acid deposition levels were high. Multiple-model evaluation frameworks as presented here allow for greater certainty in model predictions; ultimately, this type of framework should be employed when evaluating the impacts of future climate and environmental changes on soil and surface water hydrogeochemistry.