BACKGROUND: oxaliplatin with fluoropyrimidine is a "mainstay" regarding the upfront treatment of metastatic colorectal cancer (mCRC). In contrast, the efficacy and safety of oxaliplatin-based regimens in late-care settings have been poorly reported. METHODS: we identified a real-world mCRC patient cohort who were re-treated with oxaliplatin, and in which clinicopathological features were retrospectively analyzed to identify efficacy-predictive determinants (RETROX-CRC study). RESULTS: of 2606 patients, 119 fulfilled the eligibility criteria. Oxaliplatin retreatment response rate (RR) and disease control rate (DCR) were 21.6% (CI 14.4-31.0%), and 57.8% (CI 47.7-67.4). A trend towards better RR and DCR was observed among patients who had first oxaliplatin in an adjuvant setting; a poorer outcome was observed if two or more intervening treatments were delivered. Median progression-free survival (PFS) was 5.1 months (95%CI 4.3-6.1), reducing to 4.0 months (95%CI 3.07-5.13) if oxaliplatin was readministered beyond third-line (HR 2.02; 1.25-3.25; = 0.004). Safety data were retrieved in 65 patients (54.6%); 18.5% (12/65) and 7.7% (5/65) had G3-4 toxicities. Toxicities led to discontinuation in 34/119 (28.6%). CONCLUSIONS: oxaliplatin retreatment produced further RR in around one-fifth of patients and DCR 57.8%. Efficacy decreased in more pre-treated patients and around one-third of patients discontinued treatment due to adverse events. Translational studies improving patient selection are warranted.