Rivaroxaban (RXA) is a direct oral factor Xa (Xa) antagonist with a short half-life and a fast onset and offset of effect. Before elective surgery, discontinuation is recommended with an interval of at least > 24 hours. In clinical practice, this is, however, not always sufficient to achieve a residual RXA plasma concentration deemed appropriate for surgery, defined as ≤ 50 mcg/L. Our study aimed at identifying factors associated with a higher-than-expected residual RXA plasma concentration in a large group of real-life patients. This retrospective single-centre study included all patients taking RXA between 2012 and 2016 where RXA plasma concentration was determined by pharmacodynamic anti-Xa assay (518 measurements in 368 patients). Medical records were reviewed. Residual RXA plasma concentrations were then compared with expected values according to a pharmacokinetic model. Residual RXA plasma concentration was significantly higher-than-expected in patients with atrial fibrillation, impaired kidney function (glomerular filtration rate [GFR] < 60 mL/min), CYP3A4-, CYP2J2- and PGP-inhibitory co-medication including amiodarone. Impaired kidney function (odds ratio [OR], 2.22, 95% confidence interval [CI], 1.30-3.78, p = 0.003) and concomitant amiodarone intake (OR, 1.97, 95% CI, 1.04-3.72, p = 0.036) were significantly associated with RXA plasma concentrations > 50 mcg/L at 24 to 48 hours after the last RXA intake. In our group of real-life patients, impaired kidney function (GFR < 60 mL/min) and co-medication with amiodarone were independently associated with higher-than-expected residual RXA plasma concentrations. In these patients, standard intervals of RXA discontinuation may not always be sufficient before elective surgery and routine pre-operative determination of the residual RXA concentration could be advisable.