In this paper we use a simulation approach to explore the effect of variation in taxon parameters and landscape patterning on relevant source area of pollen. We use the Prentice-Sugita model, assume constant atmospheric conditions and basin morphology, and take a reductionist approach to explore the behaviour of pollen dispersal and deposition in a simple landscape scenario. Individual factors within the scenario (pollen fall speed, relative pollen productivity, size of basic unit in the landscape mosaic, patch size, rarity of individual taxa and overall number of taxa present in the landscape) are varied while all other parameters are kept constant. thus pennitting exploration of the role of different components of the system. These simulations suggest that, for basins of given size under fixed atmospheric conditions, the relevant source area of pollen is primarily an expression of the patterning of the different vegetation elements within the landscape. This has important implications for the interpretation of palaeoecological records and reconstruction of past environments. Reconstruction, especially quantitative reconstruction, requires some estimate of past relevant source area of pollen. If, as our results suggest, vegetation patterning is an important determinant of this, then it must also be taken into account when attempting to reconstruct past vegetation communities.