Forest bioenergy opportunities may be hindered by a long greenhouse gas (GHG) payback time. Estimating this payback time requires the quantification of forest-atmosphere carbon exchanges, usually through process-based simulation models. Such models are prone to large uncertainties, especially over long-term carbon fluxes from dead organic matter pools. We propose the use of whole ecosystem field-measured CO2 exchanges obtained from eddy covariance flux towers to assess the GHG mitigation potential of forest biomass projects as a way to implicitly integrate all field-level CO2 fluxes and the inter-annual variability in these fluxes. As an example, we perform the evaluation of a theoretical bioenergy project that uses tree stems as bioenergy feedstock and include multi-year measurements of net ecosystem exchange (NEE) from forest harvest chronosequences in the boreal forest of Canada to estimate the time dynamics of ecosystem CO2 exchanges following harvesting. Results from this approach are consistent with previous results using process-based models and suggest a multi-decadal payback time for our project. The time for atmospheric carbon debt repayment of bioenergy projects is highly dependent on ecosystem-level CO2 exchanges. The use of empirical NEE measurements may provide a direct evaluation of, or at least constraints on, the GHG mitigation potential of forest bioenergy projects.