Objective: To evaluate the appropriateness of the structure and the process of care for gynaecological cancer against national guidelines. Setting & Sample: Incident cases of ovarian and cervical cancer managed between I January, 1999 and 31 December, 2000 in the South East of England. Methods: Completion of an audit form, by a researcher, using the clinical record. Qualitative interviews with gynaecologists at each hospital. Main Outcome Measures: Whether hospitals meet the requirements for being a'centre'or a'unit'for gynaecological cancer and whether gynaecologists manage ovarian or cervical cancer according to the guidelines. Results: Stage was recorded in the notes of 53% (262/492) of women with ovarian cancer but information to assist in determin­ing stage was recorded in 86%. Of women who had a stage recorded by the consultant, who should have had a laparotomy and be referred for chemotherapy, 81% (169/209) were managed according to the guidelines, ignoring whether there was removal of disease to nodules of less than 1 cm. Stage was recorded in the notes of 60% (188/315) of women with cervical cancer, although information relevant to staging the patient was recorded in 66%. Overall 76% (142/188) of women with stage recorded in the notes were managed according to the guidelines but 40% (127/315) of women audited were not assessable. Only one site met the criteria for being a cancer centre. Of the remaining sites, most striking was the inability to assemble a team of professionals, including a radiologist, a pathologist and an oncology nurse to discuss individual patient care Conclusions: Retrospective note audits will be hampered by missing data in the clinical record until systems are developed to improve the quality of notes. This research suggests that guidelines based on FIGO staging for the management of ovarian and cer­vical cancer are not followed and that this may be associated with incomplete cancer management teams.