There is strong evidence that wood-based products are typically associated with lower fossil-based emissions over their lifecycle than functionally equivalent products made from other materials. However, the potential impact of large-scale material substitution at the market level remains challenging to quantify and is subject to assumptions and system boundary considerations. This paper presents a systematic review covering 44 peer-reviewed studies that quantify the substitution impacts of wood use at the level of a region or sector, to assess the commonalities and differences in scopes, system boundaries and key assumptions. We estimated the average and range of market-level substitution impacts and identify the caveats and knowledge gaps for such assessments. The results indicate an average substitution factor of 0.55 tonnes of fossil C avoided per tonne of C contained in wood harvested, with a range of 0.27–1.16 tC/tC for baseline scenarios covering all wood flows. This value depicts the average efficiency of avoided fossil emissions per unit of wood used for a certain wood use structure based on published studies but is of limited practical use as it is strictly context specific. A direct comparison between studies is complicated because a notable proportion of the studies provided insufficient information to estimate substitution factors or were not transparent in their assumptions, such as specifying which wood product is assumed to substitute for which non-wood product. A growing number of studies focus on policy-relevant analyses of the climate change mitigation potential associated with marginal changes in wood use, but market dynamics are generally considered to a limited extent. To further support decision-making, future studies could focus on changes in those end uses where increased substitution impacts could realistically be expected, while considering the various market dynamics and uncertainties.