To accurately evaluate the herbicidal effect of the mixture of saflufenacil and glyphosate, determine its optimal mixing ratio and applicable application conditions, 2 methods including indoor pot cultivation and field experiments were adopted, and a variety of Poaceae weeds (such as Digitaria sanguinalis), broad-leaved weeds (such as Conyza canadensis), and Cyperaceae weeds (Cyperus rotundus) were selected as the research objects. With clear water as the control (CK), two experimental methods were adopted: indoor pot experiments and field experiments. For the indoor pot experiments, treatments 1-5 were 30% glyphosate SL, 70% saflufenacil WG, 30% glyphosate SL + 2% saflufenacil WG, 30% glyphosate SL + 0.67% saflufenacil WG, and 30% glyphosate SL + 0.4% saflufenacil WG, respectively. For the field experiments, treatments 1-5 of experiment Ⅰ were foliar sprays of 30% glyphosate SL, 70% saflufenacil WG, and 30% glyphosate SL + 70% saflufenacil WG diluted 6 000-fold, 7 500-fold, and 10 000-fold, respectively; based on the 5 treatments of the pot experiments, experiment Ⅱ added treatment 6: 30% glyphosate SL + 6% glufosinate-ammonium. The control effects of 30% glyphosate SL, 70% saflufenacil WG, and their mixtures with different ratios on weeds were analyzed. The results of the indoor pot experiments showed that 3 days after application, the plant control effects of treatments 3-5 on Digitaria sanguinalis and Eleusine indica were significantly higher than those of treatments 1 and 2 (P<0.05); 7 days after application, the fresh weight control effects of treatments 3-5 on Conyza canadensis and Eleusine indica were significantly higher than those of Treatments 1 and 2 (P<0.05). Among them, treatment 3 exhibited the most prominent weed control effect, with plant control efficacy ranging from 76.00% to 100% and fresh weight control efficacy from 84.49% to 96.38%. The field experiment results indicated that the plant control effects of treatments Ⅰ-3, Ⅰ-4, and Ⅰ-5 on Echinochloa crus-galli and Digitaria sanguinalis were significantly higher than those of Treatments Ⅰ-1 and Ⅰ-2 (P<0.05); Treatment Ⅰ-1 had no control effect on Conyza canadensis, while the plant control effects of treatments Ⅰ-3, Ⅰ-4, and Ⅰ-5 on this weed all reached 100%. For experiment Ⅱ, the plant control effects of treatments Ⅱ-3, Ⅱ-4, and Ⅱ-5 on Echinochloa crus-galli, Digitaria sanguinalis, Beckmannia syzigachne, Solidago canadensis, and Pterocypsela indica were significantly higher than that of treatment Ⅱ-2 (P<0.05). In conclusion, the control effects of the mixed formulations of 30% glyphosate SL and 70% saflufenacil WG with different ratios on weeds were superior to those of the single agents. This study provides a reference for the innovation of weeding technology in non-cultivated land.