Reducing greenhouse gas emissions in the built environment is key to mitigating climate change. This paper uses Life Cycle Assessment (LCA) to explore whether glue-laminated timber (glulam) has a significantly lower whole-life embodied carbon (Global Warming Potential; GWP) than functionally-equivalent structural steel. Alongside a baseline assessment that follows standard practice, a variety of further sensitivity analyses establish how differing assumptions affect results. Assessment found that the GWPs were highly dependent on: (1) the assumed end-of-life scenarios; (2) the approach used to consider biogenic CO2, and; (3) whether the timing of emissions was considered. In general, glulam had the lowest GWP when incinerated (with energy recovery) at end-of-life. However, when recycling is modelled according to current standards, glulam GWPs were in amongst those of steel, giving no clear ranking of materials. Landfilled glulam had a lower GWP than steel if some biogenic carbon was assumed to be permanently stored. That may be appropriate given recent guidance from the Intergovernmental Panel on Climate Change, but is not allowed in the updated standards for LCA of construction products (e.g. EN 15804:2019). Six additional impact categories were also assessed to give a broader environmental comparison. The relative impact of the two materials depended on the impact category assessed, with glulam generally having a similar or lower impact than steel. Given the findings of this paper, further research on end-of-life treatment and LCA methodology is critical to ensure that strategies aiming to reduce GWP by material selection are effective in practice.