To verify the reliability of transvaginal ultrasonography in diagnosing intrauterine disease and in evaluating the operability of submucous myomas and to determine the feasibility, acceptability and validity of hysteroscopy for menorrhagia, we performed a prospective 5 year study on 793 women of mean age +/- SD of 41.5 +/- 7.8 years. All the patients referred for excessive menstrual bleeding with uterine volume <12 week pregnancy who underwent complete physical examination, transvaginal ultrasonography and outpatient hysteroscopy with endometrial biopsy were included in the study. Outpatient hysteroscopy was not completed due to intolerance or was unsatisfactory due to excessive bleeding in 23 cases (2.9%). Only 28 women (3.5%) declared they would have refused the procedure had they imagined the pain involved. One case of pelvic infection was observed. Compared with hysteroscopy, transvaginal ultrasonography had 96% sensitivity, 86% specificity, 91% positive predictive value and 94% negative predictive value in the diagnosis of intrauterine abnormality. The sensitivity, specificity, positive and negative predictive values of ultrasonography in identifying submucous myomas operable hysteroscopically (intramural extension <50%) were respectively 80, 69, 83 and 65% with a k index of agreement between ultrasonography and hysteroscopy of 0.48. Thus, considering the good specificity and high negative predictive value, transvaginal ultrasonography may be suggested as the initial investigation in menorrhagic patients, limiting hysteroscopy to cases with positive or doubtful sonographic findings.