Objective: To compare the use of vaginoscopic versus traditionalhysteroscopy in the evaluation of endometrial cavity.Design:Prospective, randomized, single blinded, clinical trial(Canadian Task Force classification I).Setting: University-affiliated hospital in Hong Kong.Patients: Ninety women scheduled for diagnostic hysteroscopyunder no anaesthesia.Methods: Women were randomised to undergo eithervaginoscopic hysteroscopy using the H Pipelle for endometrialsampling (n = 45) or traditional hysteroscopy using the standardPipelle (n = 45). Both procedures were performed under noanaesthesia using a rigid 4.5 mm hysteroscope. Main outcomemeasures analyzed were pain scores using a 10-cm visual analoguescale during hysteroscopy, endometrial biopsy and the overall painsore of procedure, successfulness and duration of each procedure,and adequacy of endometrial sample obtained.Results: The success rate for vaginoscopic and traditionalhysteroscopy were 93.33% and 100%, respectively (P = 0.24).There was no significant difference in the mean pain score andprocedure duration between the two hysteroscopic approaches.Endometrial sampling with H Pipelle was significantly quicker byabout 45 s compared to the standard Pipelle [duration(mean ± SD, min): 1.46 ± 0.72 vs. 2.20 ± 1.19 respectively,P = 0.001] with similar biopsy adequacy. Most women (95.5% inboth approaches) found the procedure acceptable. There was nointra- and postoperative complications.Conclusions: Vaginoscopic and traditional hysteroscopicapproaches are similar in safety, feasibility and women’s painexperience. Shorter duration is needed to obtain an endometrialsampling using the H Pipelle than the standard Pipelle.