e15555 Background: Fruquintinib, a vascular endothelial growth factor receptor (VEGFR) inhibitor, is a new targeting anti-cancer agent for refractory metastatic colorectal cancer (mCRC). Promising antitumor effect of fruquintinib monotherapy was verified in a phase III FRESCO randomized study of mCRC patients, and only 30% patients were prior bevacizumab-treated. However, comprehensive effect evaluation of fruquintinib in mCRC patients who have previously failed in bevacizumab therapy were scarce. Methods: Forty patients with metastatic CRC who had previously failed at least 2 lines of standard regimen containing bevacizumab in Longhua Hospital affiliated to Shanghai University of Traditional Chinese Medicine from March 2019 to February 2020 were enrolled. These patients were administrated with fruquintinib (5mg orally, once daily) in cycles of 3 weeks on/ 1 week off. Furthermore, the primary endpoint was overall survival (OS). Secondary endpoints mainly included progression-free survival (PFS), disease control rate (DCR), objective response rate (ORR), and safety profile. Statistical analyses were performed using IBM SPSS statistics software and ‘R’ statistical software with the ‘forest plot’ package. Subsequently, the variables were screened and verified through LASSO regression with the R software “glmnet” package. Results: The mOS was 8.1months (95%CI, 5.6667-12.8667), mPFS was 4.1833 months (95% CI, 3.5333-5.0667), and DCR was 65% (26 SD in 40 patients). Particularly, gender (HR: 7.11, 95%CI:1.86-27.16,P = 0.005), age (HR:0.91, 95%CI:0.85-0.98,P = 0.012), ECOG(HR:3.75, 95%CI:1.13-12.42,P = 0.03), medication duration (HR:0.62, 95%CI:0.48-0.81,P<0.001),and the number of metastases(HR:2.95,95%CI:1.21-7.18,P = 0.017) identified the statistically significant association with OS. Eliminating the influence of time factor, ECOG and medication cycle obtained by Lasso regression were the risk factors for OS. The most frequently reported adverse events were mainly hypertension, proteinuria, hand-foot skin reaction, fatigue and diarrhea, and any new safety signals were not observed. Conclusions: Fruquintinib showed an acceptable safety profile and promising efficacy in patients with metastatic colorectal cancer failed in previously bevacizumab-based therapy. Future studies of fruquintinib should focus on identifying the patients most likely to benefit and on minimizing toxicity.