Currently, wood and wood construction materials have limitations in how carbon fluxes are accounted for in life cycle assessments. The biogenic carbon balance of wood is often considered to be neutral, meaning that the carbon sequestered by biomass through photosynthesis is considered equal to the carbon feedstock in wood that is eventually released throughout its life cycle. Several publications have recently shown that this assumption could lead to accounting errors. This research work aims to improve the biogenic carbon accounting of the forestry phase of the life cycle of softwood products. This involved specifically modelling carbon fluxes as a function of tree species, growing conditions and forest management practices, from Canadian managed forests. A baseline natural forest scenario was run for 1000 years until the carbon stocks were assumed to reach an approximate steady-state, followed immediately by a harvest scenario that was simulated for another 100 years. The ecosystem carbon costs of the harvest activity were calculated for 117 species and region forest landscapes across Canada and expressed per cubic meter of harvested wood. Most landscapes showed net sequestration after 100 years of harvest history. Exceptions to this included outlier landscapes characterized by low average annual temperatures and precipitation where slightly positive values (net emissions) were found. The mean time to ecosystem cost neutrality for each species ranged from 16 to 60 years. Knowing the time since forest management has started on a particular forest landscape now enables managers to obtain an estimate of ecosystem carbon cost per cubic meter of wood harvested for most of Canada's forests and commercial tree species. These ecosystem carbon costs can be used to generate regionalized cradle-to-gate life cycle inventories for harvested wood products across Canada.