Local and regional distributions of Thymus serpyllum, in the southeastern part of Sweden, were examined by combining experimental studies of recruitment limitation and germination, and descriptive studies of distribution range in relation to habitat and management history. The spatial pattern at a regional scale in the county of Sodermanland was examined in relation to semi-natural pastures and graves from the Iron Age. The distribution of T. serpyllum was also examined in two parishes. Thymus serpyllum was exclusively found in managed, unfertilized areas, mainly semi-natural pastures and to some extent in road verges, and almost always in dry parts. Populations of T. serpyllum were found in 32 of the 42 pastures that contained Iron Age graves, but only in 5 of 42 pastures without graves. The distribution of T. serpyllum was also more or less congruent with the distribution of Iron Age graves, in both parishes. The seed sowing experiment showed that the germination rate, winter survival and recruitment were significantly higher in disturbed (removal of ground cover) plots, for both dry and mesic vegetation. Since the establishment of T. serpyllum occurred both in dry and mesic parts of semi-natural pastures, whereas T. serpyllum is confined to dry parts, the local limitation of distribution may be due to poor dispersal or due to effects acting on later life cycle stages. The results showed that the regional distribution of T. serpyllum is likely to be dispersal limited recruitment after seed sowing was equally good al sites with or without established populations. Seeds survived and germinated after heat treatment, with temperatures at 60 degrees C, 80 degrees C and 100 degrees C, but germination rates were not higher than in the control. The results of this study indicate that the distribution of T. serpyllum is dependent on human activities, both for dispersal and for the maintenance of established populations. Long, continuous management by grazing is obviously important for the maintenance of these populations. The association with Iron Age graves may reflect either long continuity, or accidental or intentional dispersal by humans.