Greenhouse gas emissions from extracting and manufacturing building materials, often termed “embodied carbon,” are produced before buildings are occupied and are more critical to meeting global climate targets than commonly assumed. In order to motivate reductions in embodied carbon, we need better data and established benchmarks. Although Life Cycle Assessment (LCA) methods have been used to analyze individual buildings, there has not been an agreed-upon understanding of the order magnitude and range of variation of the embodied carbon of buildings. In order to address this knowledge gap, the largest known database of building embodied carbon was compiled, normalized, and analyzed. In addition to establishing the range of embodied carbon values, this research identified sources of uncertainty and proposed strategies to advance embodied carbon benchmarking practice.