BACKGROUND: Endometrial ablation is an established and effective treatment for women with heavy menstrual bleeding, but many patients require further treatment, with approximately 10% ultimately undergoing a hysterectomy within 2 years. The insertion of a levonorgestrel-releasing intrauterine system (LNG-IUS) immediately after endometrial ablation might inactivate residual untreated endometrium or prevent its regeneration, thereby preventing recurrent symptoms necessitating a hysterectomy. The aim of the study was to assess whether LNG-IUS insertion immediately after endometrial ablation in women with heavy menstrual bleeding reduced the need for subsequent hysterectomy within 2 years compared with endometrial ablation alone. METHODS: This multicentre, open-label, randomised, controlled, superiority trial was conducted in 34 hospitals in the Netherlands. Women younger than 60 years with heavy menstrual bleeding who were eligible for endometrial ablation were included and randomly assigned (1:1) to combined treatment comprising endometrial ablation plus LNG-IUS or to standard care comprising endometrial ablation alone (stratified by dysmenorrhoea and age). Masking was not applied. Endometrial ablation was done using electrical radiofrequency ablation or by transcervical resection of the endometrium. In participants allocated to combined treatment, intrauterine insertion of an LNG-IUS containing a total of 52 mg of levonorgestrel (releasing 20 μg per day in the first year) was done immediately after the ablative procedure. The primary outcome was the incidence of hysterectomy within 2 years, analysed on a modified intention-to-treat basis in all participants with available data. Serious adverse events were systematically assessed at 6 weeks, 12 months, and 24 months after the intervention. This study was registered on the International Clinical Trials Registry Platform (NL-OMON54839) and is complete. FINDINGS: Between Nov 19, 2019, and April 5, 2023, 715 participants were randomly assigned to combined treatment (n=359) or endometrial ablation alone (n=356). The median age of all participants was 45 years (IQR 40-48). 42 (6%) participants had missing data for the primary outcome at 2 years of follow-up. A hysterectomy was done in 15 (4%) of 342 participants in the combined treatment group versus 34 (10%) of 331 participants in the endometrial ablation group, corresponding to a relative risk (RR) of 0·43 (95% CI 0·24 to 0·77; p=0·0050), with a risk difference of -0·06 (95% CI -0·10 to -0·02; p=0·0040). Serious adverse events were reported in 27 (8%) of 346 participants in the combined treatment group and 26 (8%) of 327 participants in the endometrial ablation alone group (RR 0·98, 95% CI 0·59-1·65; p=0·94), and there were no treatment-related deaths. INTERPRETATION: The combination of endometrial ablation and LNG-IUS significantly reduced the incidence of hysterectomy compared with endometrial ablation alone. The combined treatment appears to be a safe adjunctive strategy for heavy menstrual bleeding that does not increase the risk of complications. FUNDING: The Netherlands Organization for Health Research and Development.