Throughout much of earth history globally warm intervals prevailed over ice age conditions. Without the constraints imposed by cold climates, forest vegetation populated high-latitude land masses. There are no modern analogs for these biomes, thus fossil remains provide information vital to understanding the structure and function of these forests. This paper reviews methods (some common in neoecology) by which forest structure, biomass and productivity can be quantified from analysis of fossil remains. Common metrics of forest structure, such as stand density, stem diameter, basal area and tree height, can be determined for most fossil forest localities using the methods described herein. When fossils are preserved as in situ forest layers (so-called fossil forests), additional analyses involving detailed stem measurement or the application of allometric regression models may yield estimates of aboveground biomass. Annual growth rates, when paired with estimates of stem wood volume production, provide an enhanced view of ecosystem energy flow. A review of published studies on polar fossil forests indicates that forest vegetation has occupied high-latitude landmasses since the Late Permian. Although these forests seem to be structurally similar to modern day forests, they differ in that they sequestered large amounts of biomass in warm polar climates. Middle Jurassic forests from New Zealand and early Cenozoic Metasequoia-dominated forests from the Canadian Arctic seem to have sequestered similar amounts of biomass despite differences in plant community taxonomic composition. The lack of dramatic differences between fossil and modern forest ecosystem structure and function may point to general constraints on forest biology that are supported by theoretical models.