Objective: The aim of this work was to examine three types of radical vaginal hysterectomy with different degrees of radicality, performed in order to reduce surgical complications and sequelae in different indications, and to test the feasibility of a new simple and quick technique for extraperitoneal pelvic lymphadenectomy to be used in combination with radical vaginal hysterectomy for treatment of cervical cancer. In this way the advantages of vaginal surgery (e.g.: unnecessary general anaesthesia, reduced surgical trauma, applicability to obese and poor surgical risk patients, fast time-saving procedure) can be preserved. Methods: We compared retrospectively the long-term results of radical vaginal and radical abdominal operations in a large series of stage IB-IIA cervical cancer patients treated at our institution in Florence from 1968 to 1983. Furthermore, we analysed the results of our experience from 1995 to 1998, when we performed extraperitoneal pelvic lymphadenectomy, followed by radical vaginal hysterectomy, on 48 patients affected by cervical cancer. Extraperitoneal pelvic lymphadenectomy was performed through two small abdo­minal incisions (6-7 cm). Twenty-two patients (45%) were obese (BMI>30 kg/m²) and 20 were poor surgical risks. FIGO stage was IBl in 18 cases, IB2 in eight, IIA in six, IIB in 12, IIIB in four. Neoadjuvant chemotherapy was given in 12 cases and preoperative irradiation was given in ten. General and regional anaesthesia were used in 30 (62.5%) and in 18 (37.5%) cases, respectively. Results: As for past experience, in stage IB the five-year survival of 356 patients who underwent radical vaginal hysterectomy and that of 288 who had radical abdominal hysterectomy with pelvic lymphadenectomy were 81 % and 75%, respectively (p < 0.05). Surgical complications were fewer with no mortality in the first group. In stage HA, survival rates were 68% for radical vaginal hysterectomy and 64% for radical abdominal hysterectomy, in 76 and 64 cases, respectively (p = n. s.). As for the more recent expe­rience, median operative time for extraperitoneal pelvic lymphadenectomy was 20 minutes for each side (range 15-36). In each patient a median of 26 lymph nodes were removed (range 16-48). Positive nodes were found in 12 cases (25%). Median operative time for radical vaginal hysterectomy was 40 minutes (range 30-65). Extraperitoneal pelvic lymphadenectomy complications included: lymphocyst, five cases (10%) and retroperitoneal hematoma, one (2%); all occurred at the beginning of the experience. Radical vaginal hysterectomy complications included: ureteral stenosis, one (2%) and uretero-vaginal fistula, one (2%). All complications occurred in patients who received radiotherapy or chemotherapy preoperatively. Median hospital stay was ten days (range 6-20). Conclusions: The results of our work demonstrate that our technique for extraperitoneal pelvic lymphadenectomy shows a good applicability to cervical cancer patients submitted to radical vaginal hysterectomy, which has a high rate of cure for stage IB and IIA as shown by our past experience. The procedure of extraperitoneal pelvic lymphadenectomy was quick, easy, and safe, and its realizanon was not detrimental to the advantages of radical vaginal hysterectomy. Our experience supports the continued use of his combined extraperitoneal and vaginal approach in the treatment of cervical cancer. Moreover, the three classes of radical vaginal hysterectomy allow tailoring the type of the operation to the clinical and physical characteristics of the patients.