We write this commentary in response to a recent paper by Barros et al. (2020) which reports rates of N2 fixation of 0–265 kg ha-1 (above-ground biomass only with no accounting of fixed N in the roots) by 15 sorghum (Sorghum bicolor (L.) Moench) genotypes at four sites in Brazil. The high values reported require careful consideration because they are up to ten times higher than observed elsewhere in Brazil on sorghum inoculated with diazotrophs specifically selected from parallel glasshouse experiments (dos Santos et al. 2017). Sorghum is a widely-grown poaceous grain crop, but one not known for its N2-fixing capability. Based on the data presented, we calculate the rate of N2 fixation would have needed to exceed 3 kg ha-1 day-1 during the first 90 days of growth for the highest fixing sorghum genotype, a rate higher than observed for most N2-fixing legumes (e.g. soybean, Glycine max (L.) Merrill; La Menza et al. 2020). Barros et al. (2020) used a methodology that relies on small natural differences in 15 N abundance (usually expressed as δ15N; parts per thousand relative to atmospheric N2) which generally occur between atmospheric N2 and plant-available N in the soil, the latter sampled with a known non-N2-fixing plant. The N2-fixing plants, partly or wholly rely on atmospheric N2 as an N source for growth, display significantly lower δ15N than plants with lesser or no dependence on N2 fixation. At the two sites at which sorghum δ15N were significantly different to the δ15N of the non-N2-fixing reference plants, the median Ndfa was calculated to be 52% and shoot N fixed 45 kg N ha-1.