Charcoal preserved in lake sediments, peat, and soils provides a record of past fire occurrence. An understanding of fire history is important in evaluating interactions between vegetation, climate and human disturbances through at least the last several millennia. In this paper we review information concerning the production, dispersal, sedimentation and preservation of charcoal. We present examples of studies that have used charcoal analysis in palaeoecological reconstructions, with special emphasis on analytical techniques and problems of interpretation. Unlike pollen, which is produced continuously in fairly constant amounts, charcoal is produced in large quantities but at irregular intervals. These are a function of fire regimes that are often unique to specific vegetation types and/or climatic regions. Charcoal particles vary in size from sub-microscopic to macroscopic, with small particles presumably being transported further by wind and water than large particles. Charcoal preserves well, but it may be subject to breakage, especially when transported by water. We present theoretical models of dispersal and discuss potential problems associated with post-depositional mixing. A variety of charcoal analysis techniques have been employed during the past four decades. Most involve microscopic identification and quantification of numbers or size of individual fragments occurring in samples prepared for pollen analysis. The most commonly used method - estimating charcoal area by categorizing particles in several size classes - is both tedious and time consuming, and recently introduced techniques attempt to estimate past fire occurrence based upon point count estimation, elemental carbon analysis, magnetic measurement of sediments, electron microscope, and spectrographic analyses. A lack of standardization both within and among analysis techniques has hampered interpretation of charcoal profiles. Taphonomic processes affecting charcoal are less well understood than for pollen, and as a result interpretations of historic interactions between vegetation and fire based upon pollen and charcoal analyses are difficult. We review several studies through which advances have been made and suggest questions for future study.