A study consisting of 13 field experiments was conducted during 2014–2016 in southwestern Ontario and southcentral Nebraska (Clay Center) to determine the effect of late-emerging weeds on the yield of glyphosate-resistant soybean. Soybean was maintained weed-free with glyphosate (900 g ae ha^-1) up to the VC (cotyledon), V1 (first trifoliate), V2 (second trifoliate), V3 (third trifoliate), V4 (fourth trifoliate), and R1 (beginning of flowering) growth stages, after which weeds were allowed to naturally infest the soybean plots. The total weed density was reduced to 24%, 63%, 67%, 72%, 76%, and 92% in Environment 1 (Exeter, Harrow, and Ridgetown) when soybean was maintained weed-free up to the VC, V1, V2, V3, V4, and R1 soybean growth stages, respectively. The total weed biomass was reduced by 33%, 82%, 95%, 97%, 97%, and 100% in Environment 1 (Exeter, Harrow, and Ridgetown) and 28%, 100%, 100%, 100%, 100%, and 100% in Environment 2 (Clay Center) when soybean was maintained weed-free up to the VC, V1, V2, V3, V4, and R1 stages, respectively. The critical weed-free periods for a 2.5%, 5%, and 10% yield loss in soybean were the V1–V2, VC–V1, and VC–V1 soybean stages in Environment 1 (Exeter, Harrow, and Ridgetown) and V2–V3, V2–V3, and V1–V2 soybean stages in Environment 2 (Clay Center), respectively. For the weed species evaluated, there was a minimal reduction in weed biomass (5% or less) when soybean was maintained weed-free beyond the V3 soybean growth stage. These results shows that soybean must be maintained weed-free up to the V3 growth stage to minimize yield loss due to weed interference.