BACKGROUND: Previous trials from our group suggested an overall survival benefit with five cycles of adjuvant full-dose epirubicin plus ifosfamide in localised high-risk soft-tissue sarcoma of the extremities or trunk wall, and no difference in overall survival benefit between three cycles versus five cycles of the same neoadjuvant regimen. We aimed to show the superiority of the neoadjuvant administration of histotype-tailored regimen to standard chemotherapy. METHODS: For this international, open-label, randomised, controlled, phase 3, multicentre trial, patients were enrolled from 32 hospitals in Italy, Spain, France, and Poland. Eligible patients were aged 18 years or older with localised, high-risk (high malignancy grade, 5 cm or longer in diameter, and deeply located according to the investing fascia), soft-tissue sarcoma of the extremities or trunk wall and belonging to one of five histological subtypes: high-grade myxoid liposarcoma, leiomyosarcoma, synovial sarcoma, malignant peripheral nerve sheath tumour, and undifferentiated pleomorphic sarcoma. Patients were randomly assigned (1:1) to receive three cycles of full-dose standard chemotherapy (epirubicin 60 mg/m2 per day [short infusion, days 1 and 2] plus ifosfamide 3 g/m2 per day [days 1, 2, and 3], repeated every 21 days) or histotype-tailored chemotherapy: for high-grade myxoid liposarcoma, trabectedin 1.3 mg/m2 via 24-h continuous infusion, repeated every 21 days; for leiomyosarcoma, gemcitabine 1800 mg/m2 on day 1 intravenously over 180 min plus dacarbazine 500 mg/m2 on day 1 intravenously over 20 min, repeated every 14 days; for synovial sarcoma, high-dose ifosfamide 14 g/m2, given over 14 days via an external infusion pump, every 28 days; for malignant peripheral nerve sheath tumour, intravenous etoposide 150 mg/m2 per day (days 1, 2, and 3) plus intravenous ifosfamide 3 g/m2 per day (days 1, 2, and 3), repeated every 21 days; and for undifferentiated pleomorphic sarcoma, gemcitabine 900 mg/m2 on days 1 and 8 intravenously over 90 min plus docetaxel 75 mg/m2 on day 8 intravenously over 1 h, repeated every 21 days. Randomisation was stratified by administration of preoperative radiotherapy and by country of enrolment. Computer-generated random lists were prepared by use of permuted balanced blocks of size 4 and 6 in random sequence. An internet-based randomisation system ensured concealment of the treatment assignment until the patient had been registered into the system. No masking of treatment assignments was done. The primary endpoint was disease-free survival. The primary and safety analyses were planned in the intention-to-treat population. We did yearly futility analyses on an intention-to-treat basis. The study was registered with ClinicalTrials.gov, number NCT01710176, and with the European Union Drug Regulating Authorities Clinical Trials, number EUDRACT 2010-023484-17, and is closed to patient entry. FINDINGS: Between May 19, 2011, and May 13, 2016, 287 patients were randomly assigned to a group (145 to standard chemotherapy and 142 to histotype-tailored chemotherapy), all of whom, except one patient assigned to standard chemotherapy, were included in the efficacy analysis (97 [34%] with undifferentiated pleomorphic sarcoma; 64 [22%] with high-grade myxoid liposarcoma; 70 [24%] with synovial sarcoma; 27 [9%] with malignant peripheral nerve sheath tumour; and 28 [10%] with leiomyosarcoma). At the third futility analysis, with a median follow-up of 12.3 months (IQR 2.75-28.20), the projected disease-free survival at 46 months was 62% (95% CI 48-77) in the standard chemotherapy group and 38% (22-55) in the histotype-tailored chemotherapy group (stratified log-rank p=0.004; hazard ratio 2.00, 95% CI 1.22-3.26; p=0.006). The most common grade 3 or higher adverse events in the standard chemotherapy group (n=125) were neutropenia (107 [86%]), anaemia (24 [19%]), and thrombocytopenia (21 [17%]); the most common grade 3 or higher adverse event in the histotype-tailored chemotherapy group (n=114) was neutropenia (30 [26%]). No treatment-related deaths were reported in both groups. In agreement with the Independent Data Monitoring Committee, the study was closed to patient entry after the third futility analysis. INTERPRETATION: In a population of patients with high-risk soft-tissue sarcoma, we did not show any benefit of a neoadjuvant histotype-tailored chemotherapy regimen over the standard chemotherapy regimen. The benefit seen with the standard chemotherapy regimen suggests that this benefit might be the added value of neoadjuvant chemotherapy itself in patients with high-risk soft-tissue sarcoma. FUNDING: European Union grant (Eurosarc FP7 278472).