Study design: Secondary analysis of multicentre prospective observational data. Objectives: To describe methods of bladder emptying at admission and discharge in patients with recently acquired spinal cord injury (SCI) and to describe predictors of bladder emptying methods at discharge. Setting: First inpatient rehabilitation in specialised rehabilitation centres in the Netherlands. Methods: Data from the Dutch Spinal Cord Injury Database collected between 2015 and 2019 were used. McNemar-Bowker test was used to evaluate if bladder emptying methods differed over time; One-Way ANOVA and Chi-Square tests to see if bladder emptying methods differed by demographic and injury-related characteristics. Binary logistic regression was used to predict the type of bladder emptying at discharge with demographic and injury-related characteristics measured at admission. Results: Of 1403 patients, 44.1% had cervical, 38.4% thoracic and 17.5% lumbosacral lesions at admission. AIS classification was mostly D (63.8%). The method of bladder emptying changed significantly (p < 0.001) from admission to discharge: decrease of clean intermittent assisted catheterisation (17.1% to 4.1%) and indwelling catheter (33.4% to 16.3%) and increase in clean intermittent self-catheterisation (CISC, 7.8% to 22.2%) and normal voiding (40.2% to 56.1%). Age, sex, SCI level, AIS classification and level of independence predicted the method of bladder emptying at discharge (all p < 0.001). Conclusions: During first inpatient rehabilitation, the method of bladder emptying changed resulting in more patients discharged with normal voiding and CISC. Age, sex, SCI level, AIS classification and level of independence in self-care were all confirmed as factors playing a role in this change.