To examine the relationship between plant presence and pollen presence, modern pollen samples were collected and counted from natural moss polsters in the UK uplands. The herbaceous pollen signal is difficult to interpret in paleoecological analysis as pollen presence is often rare for individual herb taxa. Indices of association have previously been used to determine association from over- and under-representation; however, previous surveys have used spatially limited single-source areas. Indices of Association (IOA) between pollen and plant presence at increasing distance from the pollen source, to 50 m, are presented in a new spatially extended IOA (SEIOA) method. For several pollen taxa, including Plantago lanceolata, Ranunculus acris-type, Rumex and the Cichorioideae family, association increased progressively, when plants at greater distance were included in the calculation, and over-representation was reduced. Other taxa become increasingly under-represented at the landscape scale, e.g., Filipendula, Apiaceae, Cerastium indicating true ultra-local source areas. A weak association and increasing under-representation indices, cautions that pollen absence of these types in paleoecological analysis is not a secure indicator of plant absence. Extended IOA values can aid the interpretation of the rare-type pollen spectra in paleoecological analyses. The method shows potential for developing individual pollen source areas (iPSAs) for herbaceous pollen types. These iPSAs together with the probability of correct application, determined by accuracy and association scores, are useful when applying pollen correction factors in pollen-derived vegetation community and landscape models. Further indices and significance testing are suggested to enhance understanding of the association and the rate of correct classification. The results can assist both interpretation of community composition from the herbaceous pollen spectra in local-scale paleoecological studies, and understanding the limits to this interpretation in paleoecological analysis.