Objective: To explore the related factors and to construct the nomogram model and artificial neural network model (ANN)for prediction of the neoadjuvant chemotherapy sensitivity in cervical cancer.Methods: Clinical data of 468 patients with local advanced cervical cancer who received neoadjuvant chemotherapy in Sun-Yat-Sen University Cancer Center from January 2005 to December 2013 were retrospectively analyzed.The patients who were diagnosed during 2005-2011 were assigned to model-development cohorts (n=405) and the others were assigned to validation cohorts (n=63).A multivariate logistic model was created to analyze the clinical data before chemotherapy,from which a nomogram model and the ANN model were developed and validated.Results: In this study,family history of cancer,pre-treatment level of squamous cell carcinoma antigen (SCC),histological type,chemotherapy cycle and tumor size assessed by radiography were found to be independent predictors for neoadjuvant chemotherapy sensitivity in cervical cancer.The nomogram incorporating these five predictors demonstrated intermediate discrimination and calibration (concordance index=0.6733;,95% confidence interval (CI),0.6047-0.7418).In the validation cohort,the discrimination accuracy was 0.7572 (95%CI;0.6184-0.896).The artificial neural network model was successfully built with the independent variances of family history,pre-treatment SCC,pre-treatment hemoglobin,histological type,chemotherapy cycle and tumor size by radiographyThe AUC of predicting chemotherapy effectiveness was 81.7% with 0.7397 of the best prediction performance truncation therapies.Conclusion: A robust nomogram and ANN models predicting neoadjuvant chemotherapy sensitivity were developed.These models may improve clinical trial design and help physicians to decide whether neoadjuvant chemotherapy should be performed.