MacNaughton et al. recently published an article entitled, "Energy savings, emissions reductions, and health co-benefits of the green building movement" in which they claim to calculate the environmental co-benefits associated with the (assumed) reduced energy use of green buildings. They consider only LEED (Leadership in Energy and Environmental Design) commercial buildings and make two fundamental assumptions: (1) that each LEED building, year after year, achieves the energy savings projected by its design team, and (2) that the fuel mix of LEED buildings is the same as the average mix for other buildings in the same geographic region.Here we show that these assumptions are not supported by data. Numerous studies have shown that buildings, on average, use significantly more energy than projected by design simulations. Furthermore, a decade of research suggests that LEED-certified buildings, on average, achieve little or no primary energy savings relative to other similar buildings. In addition, evidence suggests that any reduction in site energy is typically achieved through increased electric use and corresponding off-site energy loss. The environmental benefits of LEED buildings calculated by MacNaughton et al. have dubious value because they are based on assumptions that are inconsistent with measured LEED building energy performance.