Backward Lagrangian stochastic (bLS) modelling is widely used to assess emission rates from agricultural activities. A large number of investigations using bLS modelling estimates ammonia (NH3) emissions from confined areas. NH3 is known to efficiently deposit on surfaces, especially if they are moist. Most bLS models do not include deposition mechanisms. Three release experiments using an artificial source were conducted. A gas mixture of 5% NH3 and 95% methane (CH4) was released. Lineintegrated measurements of NH3 and CH4 downwind of the source were carried out and recovery rates were calculated using bLS dispersion modelling. The recovery rates averaged to 106% for CH4 and 84% for NH3, respectively. The comparison between NH3 and the inert trace gas CH4 was used to assess the effect of (dry) deposition. The bLS model was extended by a dry deposition algorithm. Modelled concentration reductions due to the inclusion of dry deposition were comparable to the observed differences in the recovery rates of the inert CH4 and the deposition affected NH3.