65 Background: The efficacy of late line treatments for metastatic colorectal cancer (mCRC) is still limited. Antiangiogenic therapy is one of standard treatments in mCRC. Anlotinib is a multitarget tyrosine kinase inhibitor, mainly targets VEGFR1-3. This phase III trial was conducted to evaluate anti-tumor efficacy and safety of anlotinib monotherapy in refractory mCRC. Methods: Aged 18 years or older mCRC patients were enrolled in 33 Chinese hospitals. They had received at least 2 lines of chemotherapy based treatments; had ECOG PS 0-1and adequate bone marrow, liver and renal function. Enrolled patients were 2-to-1 randomly assigned into anlotinib or placebo group and stratified by previous VEGF-targeting therapy (y/n) and time from diagnosis to metastases (</≥18 months). Anlotinib (12 mg/d, oral, d1-14, q3w) or matched placebo was administered plus best supportive care. Primary end point was Overall Survival (OS). Secondary endpoints included Progression Free Survival (PFS), Objective Response Rate (ORR), Disease Control Rate (DCR), safety and quality of life. Results: Between December 2014 and August 2016, 421 patients were enrolled (282 in anlotinib and 137 in placebo group). By the cut-off date (August 31, 2018), the median OS was 8.6 months (95% CI 7.8-9.7) in anlotinib and 7.2 months (6.2-8.8) in placebo without statistical difference (p=0.87; HR=1.02 [0.82-1.27]). The median PFS in anlotinib group (4.1 months; 3.4-4.5) was significantly improved over that in placebo (1.5 months; 1.4-1.5), with a p<0.0001 and a HR of 0.34 (0.27-0.43). Improvements of ORR and DCR were observed in anlotinib over placebo (Table). In subgroup patients with RAS/KRAS/BRAF wildtype, the median OS was statistically prolonged in anlotinib (11.0 months; 8.6-14.1; n=103) than that in placebo (6.7 months; 3.5- 11.1; n=46), HR=0.68 (0.47, 0.99), p=0.043. 52.5% of patients in anlotinib and 19.7% in placebo group experienced grade 3 or above treatment related adverse events (TRAEs). The most common grade ≥3 anlotinib related adverse events were hypertension (20.9%), increased γ-GT (7.1%) and hand-foot syndrome (6.4%), which were recovered by reducing dosage. Only 2 patients in anlotinib group had grade 5 TRAE, which were both hemorrhage. Conclusions: Anlotinib could provide clinical benefits by substantially prolonged PFS with manageable toxicity for Chinese refractory mCRC patients. Although median OS did not reach significant improvement. Those with RAS/KRAS/BRAF wildtype might be the potential patients in anlotinib treatment. Clinical trial information: NCT02332499. [Table: see text]