The interpretation of palynological data into a history of the plant communities in a region requires (1) the estimation of the pollen density in the samples, (2) the knowledge of the relative representation rates of the vegetation elements in the pollen spectra and (3) a careful distinction between air-borne and water-borne pollen because of their different recording values. In the present article, a method is described for pollen density determination which is convenient when cores have to be analysed consisting of peat, gyttja and clastic sediments. This method involves the addition of a foreign pollen. Examples are given of the variation which may occur in the pollen density of sediments in a horizontal series of samples. In three cores the pollen densities were determined and the use of this additional information for the interpretation of the analysis data is demonstrated. In a number of nature reserves a comparison was made between the composition of the local stand of vegetation and the pollen sedimentation. To this purpose moss samples were analysed. The representation of the plant communities was generally found to be fair. The dispersal of water-borne pollen was studied by a comparison of the regional vegetation types with the pollen precipitated on river banks and tidal flats; there was also made a series of analyses of the pollen content of the water, which shows that the seasonal variations are so small that the pollen which is precipitated from the water in the regions investigated is not representative for the regional vegetation type. The pollen which is deposited in such conditions may originate from three different sources: (1) the local or regional stand of vegetation, (2) forests or other plant communities in remote upstream areas and (3) peat or gyttja layers or other pollen containing deposits which are eroded by the water currents. A number of examples of these dispersal mechanisms is described in this article. In a few cases it is demonstrated how the determination of pollen density may contribute to the solution of the problem, in which proportion these three dispersal mechanisms have been responsable for the presence of the pollen in the sediment under investigation.