We present the major results from studies of fire history over the last 11000 years (Holocene) in southern Sweden, on the basis of palaeoecological analyses of peat sequences from three small peat bogs. The main objective is to emphasize the value of multiple, continuous sedimentary records of macroscopic charcoal (macro-C) for the reconstruction of local to regional past changes in fire regimes, the importance of multi-proxy studies, and the advantage of model-based estimates of plant cover from pollen data to assess the role of tree composition and human impact in fire history. The chronologies at the three study sites are based on a large number of C-14 dates from terrestrial plant remains and age-depth models are achieved using Bayesian statistics. Fire history is inferred from continuous records of macro-C and microscopic charcoal counts on pollen slides. The Landscape Reconstruction Algorithm (LRA) for pollen-based quantitative reconstruction of local vegetation cover is applied on the three pollen records for plant cover reconstruction over the entire Holocene. The results are as follows: (1) the long-term trends in fire regimes are similar between sites, i.e., frequent fires during the early Holocene until ca. 9 ka BP, low fire frequency during the mid-Holocene, and higher fire frequency from ca. 2.5 ka BP; (2) this broad trend agrees with the overall fire history of northwestern and western Europe north of the Mediterranean area, and is due to climate forcing in the early and mid-Holocene, and to anthropogenic land-use in the late Holocene; (3) the LRA estimates of plant cover at the three sites demonstrate that the relative abundance of pine played a primordial role in the early and mid-Holocene fire history; and (4) the between-site differences in the charcoal records and inferred fire history are due to local factors (i.e., relative abundance of pine, geomorphological setting, and anthropogenic land-use) and taphonomy of charcoal deposition in the small peat bogs. It is shown that continuous macro-C records are most useful to disentangle local from regional-subcontinental fire history, and climate-induced from human-induced fire regimes, and that pollen-based LRA estimates of local plant cover are more adequate than pollen percentages for the assessment of the role of plant composition on fire history.