Abstract This thesis deals with the Holocene development of palsa mires in continental Europe, especially permafrost dynamics and its consequences on vegetation succession and peat accumulation. Peat deposits of four permafrost mires in boreal and subarctic northeastern European Russia and in northern oroboreal Finland have been studied using plant macrofossil analysis, (AMS) radiocarbon dating, dry bulk density and carbon content measurements. In addition, preliminary results are available from another palsa mire in northeastern European Russia. Modern vegetation has been investigated to support the interpretation of fossil plant assemblages. Earlier literature on vegetation, stratigraphy and dating of permafrost mires in Europe has been reviewed. The vegetation of palsa mires in general is well known. As a rule, palsas are dry ombrotrophic habitats, surrounded by wet flarks of variable trophic levels. There is a lack of information about vegetation in different small-scale habitats within palsa mires, which would have been useful when studying the permafrost-vegetation relationship. Although no functional indicator species of permafrost have been found, permafrost dynamics in peat stratigraphy can often be detected with high degree of probability based on changes in vegetation. Some plant assemblages and vegetation successions are typical on permafrost, while many species rarely grow on or near to permafrost. Relatively sudden changes between dry and wet mire environments and continuously dynamic conditions are good signs of permafrost impact. Also gradual changes towards drier conditions may be caused by permafrost; in these cases the timing of first permafrost aggradation is more difficult to ascertain and can usually be pronounced only in terms of maximum and minimum ages. Changes in peat accumulation rates and even hiatuses in stratigraphy are additional tools to support the interpretation on permafrost history at the studied sites. Dry organic matter and carbon accumulation rates for different developmental stages are calculated for the five studied mires. From earlier studies this information is not available. Accumulation rates in the permafrost environment are very variable: from zero or negative rates in old palsas to as high as 100 gC/m2yr in incipient palsas. On moist plateau palsas, permafrost flarks and in unstable permafrost conditions, accumulation continues at low to moderate rates. Thermokarst processes result in decomposition of former peat deposits with important consequences for the ecosystem carbon balance, especially in plateau palsa mires. Radiocarbon datings are available from 27 permafrost mires in continental Europe; only 5 of these are situated in Russia. Many of the published dates cannot be considered reliable as dating permafrost aggradation. Based on limited material, permafrost started to develop at latest about 3000 BP in mires of northern Russia and 2500 BP in Fennoscandia. Older permafrost formation is suggested for a few sites, but the evidence is insufficient to confirm this interpretation. The oldest preserved palsas are ca. 2500–2000 14C years old. Most of the modern palsas are less than 600 14C years old. Permafrost aggradation follows the major climate development in the Holocene, with formation being most active during the coldest stages. Global warming is expected to greatly affect the Arctic in the near future, which would imply significant changes in ecosystem functioning and carbon balance of permafrost mires. This study contributes to the understanding of the possible impacts of climate change on these ecosystems using paleoecological techniques.