The terrestrial biosphere is thought to be a key component in the climatic variability seen in the paleo record. It has a direct impact on surface temperature through changes in surface albedo and evapotranspiration (so called biogeophysical effects) and in addition, has an important indirect effect through changes in vegetation and soil carbon storage (biogeochemical effects) and hence modulates the concentrations of greenhouse gases in the atmosphere. The biogeochemical and biogeophysical effects generally have opposite signs meaning that the terrestrial biosphere could potentially have played only a very minor role in the dynamics of the glacial-interglacial cycles of the late Quaternary. Here we use a fully-coupled dynamic atmosphere-ocean-vegetation General Circulation Model (GCM) to generate a set of 62 simulations spanning the last 120 ka. The analysis of these simulations elucidates the relative importance of the biogeophysical versus biogeochemical terrestrial biosphere interactions with climate. We find that the biogeophysical effects of vegetation account for up to an additional −0.84 °C global mean cooling, with regional cooling as large as −5 °C, but with considerable variability across the glacial-interglacial cycle. By comparison, while opposite in sign, our model estimates of the biogeochemical impacts are substantially smaller in magnitude. Offline simulations show a maximum of +0.33 °C warming due an increase of 25 ppm above our (pre-industrial) baseline atmospheric CO<sub>2</sub> mixing ratio. In contrast to shorter (century) time-scale projections of future terrestrial biosphere response where direct and indirect responses may at times, cancel out, we find that the biogeophysical effects consistently and strongly dominate the biogeochemical effect over the inter-glacial cycle. In addition, depending on the assumptions about soil carbon under ice-sheets and sea level rise, we find a range in terrestrial carbon storage change from a reduction in LGM carbon storage of −440 PgC, to a gain of +37 PgC. We suggest that prevailing uncertainties allow for only a small net transfer of carbon between terrestrial biosphere and ocean atmosphere implying that explaining the observed CO<sub>2</sub> ice core record could be rather simpler than previously thought.