ObjectiveTo compare adnexectomy by vaginal Natural Orifice Transluminal Endoscopic Surgery (vNOTES) versus laparoscopy.DesignParallel group, 1:1 single‐centre single‐blinded randomised trial, designed as non‐inferiority study with a margin of 15%.SettingBelgian teaching hospital.PopulationNon‐pregnant non‐virgin women with an intact uterus and without obliteration of the pouch of Douglas scheduled to undergo removal of an adnexal mass assessed to be benign on ultrasound by IOTA criteria.MethodsRandomisation to laparoscopy (control group) or vNOTES (experimental group). Stratification according to adnexal size. Blinding of participants and outcome assessors by sham incisions.Main outcome measuresThe primary outcome measure was adnexectomy by the allocated technique. Secondary outcomes included duration of surgery, pain scores and analgesics used, quality of life and adverse events.ResultsWe randomly assigned 67 participants (34 to the vNOTES group and 33 to the laparoscopy group). The primary end point was always reached in both groups: there were no conversions. We performed a sensitivity analysis for the primary outcome, assuming one conversion in the vNOTES group and no conversions in the laparoscopy group: the one‐sided 95% upper limit for the differences in proportions of conversion was estimated as 13%, which is below the predefined non‐inferiority margin of 15%. The secondary outcomes demonstrated a shorter duration of surgery, lower pain scores, lower total dose of analgesics and a trend for more adverse events in the vNOTES group.ConclusionsvNOTES is non‐inferior to laparoscopy for a successful adnexectomy without conversion. vNOTES allowed shorter operating times and less postoperative pain but there was a trend for more adverse events.