ObjectiveThe efficacy of regorafenib or fruquintinib in combination with PD-1/PD-L1 inhibitors for metastatic colorectal cancer (mCRC) treatment has not been elucidated. This study aims to systematically evaluate the efficacy and safety of this combination therapy.MethodsPubMed, Embase, Cochrane Library, and Web of Science were systematically retrieved until July 24, 2024. A meta-analysis was carried out for the overall objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), overall survival (OS), and the incidence of grade 3 or higher treatment-related adverse events (AEs). Non-overlapping 95% confidence intervals (CIs) were considered statistically significant.Results26 studies encompassing 1,409 patients were analyzed. Pooled analysis revealed an ORR of 6% (95% CI: 3%-12%), a DCR of 62% (95% CI: 55%-68%), a median PFS of 3.84 months (95% CI: 3.19-4.49 months), a median OS of 13.08 months (95% CI: 10.17-16.00 months), and an incidence rate of grade 3–4 AEs of 21% (95% CI: 15%-28%). In subgroup analyses, the fruquintinib-based regimen demonstrated significantly superior efficacy compared to regorafenib-based therapy, with higher ORR (16% [95% CI: 13%-21%] vs 3% [95% CI: 1%-9%]), DCR (79% [95% CI: 72%-85%] vs 54% [95% CI: 47%-61%]), and median PFS (5.40 months [95% CI: 4.60-6.19] vs 3.00 months [95% CI: 2.47-3.52]). Median OS was numerically but not significantly longer with fruquintinib (14.35 months [95% CI: 10.68-18.02] vs 12.70 months [95% CI: 8.79-16.61]). Liver metastasis status strongly influenced outcomes, with significantly lower ORR (3% [95% CI: 1%-13%] vs 49% [95% CI: 32%-76%]) and shorter median PFS (2.37 months [95% CI: 1.77-2.96] vs 3.50 months [95% CI: 3.09-3.91]) in patients with liver involvement.ConclusionThe combination of regorafenib or fruquintinib with PD-1/PD-L1 shows moderate efficacy and acceptable safety in the treatment of mCRC. The fruquintinib-based regimen may be superior to the regorafenib-based regimen, and patients without liver metastasis may derive greater benefits. These findings offer new insights for treating mCRC, although they should be validated through large randomized controlled trials.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO, identifier CRD42024582268