© 2020 The Author(s). Background: Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) has been increasingly used over the last decade in patients with refractory cardiogenic shock. ECMO weaning can, however, be challenging and lead to circulatory failure and death. Recent data suggest a potential benefit of levosimendan for ECMO weaning. We sought to further investigate whether the use of levosimendan could decrease the rate of ECMO weaning failure in adult patients with refractory cardiogenic shock. Methods: We performed an observational single-center cohort study. All patients undergoing VA-ECMO from January 2012 to December 2018 were eligible and divided into two groups: group levosimendan and group control (without levosimendan). The primary endpoint was VA-ECMO weaning failure defined as death during VA-ECMO treatment or within 24 h after VA-ECMO removal. Secondary outcomes were mortality at day 28 and at 6 months. The two groups were compared after propensity score matching. P < 0.05 was considered statistically significant. Results: Two hundred patients were analyzed (levosimendan group: n = 53 and control group: n = 147). No significant difference was found between groups on baseline characteristics except for ECMO duration, which was longer in the levosimendan group (10.6 ± 4.8 vs. 6.5 ± 4.7 days, p < 0.001). Levosimendan administration started 6.6 ± 5.4 days on average following ECMO implantation. After matching of 48 levosimendan patients to 78 control patients, the duration of ECMO was similar in both groups. The rate of weaning failure was 29.1% and 35.4% in levosimendan and control groups, respectively (OR: 0.69, 95%CI: 0.25-1.88). No significant difference was found between groups for all secondary outcomes. Conclusion: Levosimendan did not improve the rate of successful VA-ECMO weaning in patients with refractory cardiogenic shock. Trial registration: ClinicalTrials.gov, NCT04323709.