Background: To evaluate the long-term efficacy and toxicities of three therapeutic strategies for patients with International Federation of Gynecology and Obstetrics (FIGO) stage IB2/IIA2 cervical cancer. Methods : 206 patients with stage IB2/IIA2 cervical cancer were selected in this retrospective study. The patients were classed into three groups according to the primary therapy: primary surgical treatment (PST), neoadjuvant chemotherapy followed by radical surgery (NAC+RS), and concurrent chemoradiotherapy (CCRT). We observed the inter-group differences in patient characteristics, surgery approaches, postoperative rick factors, supplemental therapies and overall survival (OS), disease-free survival (DFS) and acute and late toxicities. Results: Less patients in NAC+RS group had deep cervical stromal invasion than those in PST group (p=0.024). No differences in lymph node metastasis, intravascular tumor embolus, vaginal margin and the need of postoperative supplement treatment were observed between PST and NAC+RS ( P =0.734, P =0.704, P =0.487 and P =0.714, respectively). With a median follow-up time of 57 months, the 3-year DFS and OS in PST, NAC+RS and CCRT were 85.6%, 79.2%, 85.7% and 87.5%, 84.9%, 85.7% ( P =0.424 and P =0.856, respectively). The most frequently observed acute toxicities were hematologic side effects. No significant inter-group differences in leukopenia, neutropenia, thrombocytopenia and anemia were observed (all P >0.05). No patient experienced grade 3-4 hepatotoxicity and nephrotoxic. Late toxicity ≥grade 3 mainly consisted of lower extremity lymphedema (4/104, 3.8%), bowel obstruction (3/104, 2.9%) and thrombosis (1/104, 1.0%) in PST, lower extremity lymphedema (5/53, 9.4%) and bowel obstruction (1/53, 1.9%) in NAC+RS and radiation proctitis (4/49, 8.2%) and femoral head necrosis (1/49, 2.0%) in CCRT. No grade 5 toxicities were observed. There was no significant difference in cumulative late adverse effects rate in the three groups ( P =0.777). The patients in NAC+RS had increased hospitalization cost than that in PST ( P =0.000) and CCRT ( P =0.000) and prolonged hospitalization time than that in PST ( P =0.000) and same as that in CCRT ( P =0.07). Conclusion: Preoperative NAC decreased the incidence of deep cervical stromal invasion but had no effect on other high risk factors and the need of postoperative supplement treatment. No inter-group differences were demonstrated in 3-year DFS and OS and cumulative acute and late adverse effects. NAC+RS would need longer hospitalization time and cost more. Nevertheless, we believe it should be further explored in prospective trials.