PURPOSE: To evaluate outcome and reassess the radiation therapy options in pelvic recurrences of cervical cancer treated initially with surgery. MATERIALS AND METHODS: In 30 patients, the prognostic factors analyzed for local control included site of recurrence (central, pelvic wall), tumor size, modality of radiation therapy, and radiation dose. Mean follow-up in survivors was 111.5 months. RESULTS: Local control was attained in (a) nine of 20 patients with central recurrence and in two of 10 with pelvic wall recurrence (p = .25); (b) none of four who received less than 50 Gy, five of nine who received 50-60 Gy, and six of 17 who received greater than 60 Gy (p = .27); and (c) five of 11 with tumor smaller than 3 cm, five of nine with tumor size 3-6 cm, and one of 10 with tumor larger than 6 cm. Multivariate analysis revealed a significant benefit of local control on survival (P = .05). Median survival for patients with central recurrence was 14.5 months compared with 9 months for those with pelvic wall recurrence. CONCLUSION: Local pelvic control depends on site and size of recurrence and radiation therapy modality and dose. Appropriate choice of brachytherapy modality is important. To improve local control and survival, more aggressive treatment is indicated, but attendant higher complications may be expected.