The drug pair consisting of Sophora flavescens Aiton (Sophorae flavescentis radix, Kushen) and Coptis chinensis Franch. (Coptidis rhizoma, Huanglian), as described in Prescriptions for Universal Relief (Pujifang), is widely used to treat laxation. Matrine and berberine are the major active components of Kushen and Huanglian, respectively. These agents have shown remarkable anti-cancer and anti-inflammatory effects. A mouse model of colorectal cancer was used to determine the most effective combination of Kushen and Huanglian against anti-colorectal cancer. The results showed that the combination of Kushen and Huanglian at a 1:1 ratio exerted the best anti-colorectal cancer effect versus other ratios. Moreover, the anti-colorectal cancer effect and potential mechanism underlying the effects of matrine and berberine were evaluated by the analysis of combination treatment or monotherapy. In addition, the chemical constituents of Kushen and Huanglian were identified and quantified by liquid chromatography-tandem mass spectrometry (LC-MS/MS). A total of 67 chemical components were identified from the Kushen–Huanglian drug pair (water extraction), and the levels of matrine and berberine were 129 and 232 µg/g, respectively. Matrine and berberine reduced the growth of colorectal cancer and relieved the pathological conditions in mice. In addition, the combination of matrine and berberine displayed better anti-colorectal cancer efficacy than monotherapy. Moreover, matrine and berberine reduced the relative abundance of Bacteroidota and Campilobacterota at phylum level and that of Helicobacter, Lachnospiraceae_NK4A136_group, Candidatus_Arthromitus, norank_f_Lachnospiraceae, Rikenella, Odoribacter, Streptococcus, norank_f_Ruminococcaceae, and Anaerotruncus at the genus level. Western blotting results demonstrated that treatment with matrine and berberine decreased the protein expressions of c-MYC and RAS, whereas it increased that of sirtuin 3 (Sirt3). The findings indicated that the combination of matrine and berberine was more effective in inhibiting colorectal cancer than monotherapy. This beneficial effect might depend on the improvement of intestinal microbiota structure and regulation of the RAS/MEK/ERK-c-MYC-Sirt3 signaling axis.