Background: Currently, there is no accepted standard chemotherapy regimen for recurrent osteosarcoma. We performed a retrospective study to evaluate the effectiveness, toxicity and clinical benefit of second line therapy with Gemcitabine (G) in combination with Docetaxel (D), in patients with relapsed high-grade osteosarcoma who have been previously treated with High-dose Ifosfamide (HD-IFO) during first-line treatment (ClinicalTrials ID: NCT04651179). Methods: Patients were eligible to the analysis according to the following criteria: episode of first relapse of osteosarcoma, previous first line treatment according to poor responders’ arm of ISG-AIEOP OS 2 Protocol based on methotrexate, cisplatin, doxorubicin and HD-IFO (3gr/m2/day, day 1-5) ± Mifamurtide, presence of measurable disease according to RECIST 1.1, treatment with at least two cycles of Gemcitabine 900 mg/m2/day, days 1 and 8 and Docetaxel 75 mg/m2 day 8 every 21 days. Primary objective: overall survival (OS) at 12 months; secondary objectives: overall response rate (ORR) (complete response + partial response), progression-free survival (PFS) at 6-12 months, toxicity and quality of life. Results: 14 patients were included in the analysis. All patients were previously treated according to poor responders’ arm of ISG-AIEOP OS 2 Protocol, 7/14 (50%) of them received Mifamurtide. Most of patients (12/14, 86%) received 4 GD cycles and the total number of cycles administered was 60. Median OS was 20 months (range 11-69); OS at 12 and 24 months was 84% (95% CI 65-100) and 51% (95% CI 22-79), respectively; ORR were 35,7% after six cycles. Median PFS was 12 months (range 2-69); 6-months and 12-months PFS were 86% (95% CI 67-100) and 56% (30-83) respectively. For 8/14 (57%) patients a surgical approach was feasible after 2 and 6 GD cycles. Previous use of Mifamurtide correlate with a better 6-months PFS rate (100% vs 57%, P = 0.0011). Also, amelioration of ECOG/Lansky score showed a positive correlation with both PFS (P = 0.01) and OS (P = 0.04). No extra-hematological toxicities grade ≥3 was observed according to CTCAE v4.03 criteria. Conclusions: This trial confirms that GD combination has an anti-tumor activity for relapsed high-grade osteosarcoma with a well-tolerated toxicity profile. Furthermore, GD guarantee a surgical tumor remotion for patients with a chemo-resistant disease such as patients previously treated with a poor responders’ arm improving their OS and PFS rate. Promising results for patients previously treated with Mifamurtide need to be confirmed in future trials. Clinical trial information: NCT04651179.