OBJECTIVE To determine referral rates and intermediate and long-term outcomes for patients consulting for menstrual disorders and referred by their general practitioner to gynaecology outpatient clinics. METHODS General practitioners' records of referrals to outpatient clinics and retrospective audit of general practice notes to determine outcomes. METHODS General practices in the Oxford Regional Health Authority area referring to 19 gynaecology outpatient clinics. METHODS 205 patients aged 15-59, referred in 1983/4 and follow up in 1988/9. METHODS Immediate outcomes: the initiation by hospital specialists of investigation, treatment or advice. Five year outcomes: general practice consultation rates and symptom prevalence. RESULTS Of 18,754 index referrals recorded by 33 practices over a period of 6 months, 2513 (13%) went to gynaecology clinics. Menstrual disorders constituted 21% (n = 539) of the gynaecology referrals; there was more than three-fold variation between the practices in referral rates. In the 5 years following the index referral, of the 205 audited patients 167 (81%) had been admitted to hospital, 91 (44%) had had a hysterectomy (including 87 (60%) of the 145 patients referred for menorrhagia), 98 (48%) had dilatation and curettage; 25 (12%) received only drug therapy; and 10 (5%) had no active treatment for these symptoms from either the specialist or the general practitioner. Only 29 (14%) had consulted their general practitioners about menstrual problems in the 12 months preceding the audit. CONCLUSIONS Guidelines are needed to assist referral decision-making. If audit is to be used to promote good practice these guidelines should consider the patients' anxieties and preferences, as well as the most appropriate use of investigations and treatments.