Many analyses of cancer survival data prefer to use Cox proportional hazards (CPH) model which had influence on development of cancer covariates in the field of survival analysis. Study attempted to evaluate parametric, semi-parametric and non parametric approach to find a model for available cervical cancer survival data from ORCI in order to show applicability and workability in medical data. Life table method calculates probabilities of censored patients when survival times grouped and number of patients in every interval for 161 female patients diagnosed with cervical cancer and treated at ORCI between 2014 and 2015and findings shows that the survival of patients was poor with probability of survival was 0.194 and patients with latter cancer stage such as stage had an increased risk of death. Kaplan-Meier product limit approach calculate survival times of patients easily without considering the effect of covariates and understandable. The logistic regression analysis and Cox regression model with Breslow method determined significant covariates that affect the survival times as menopause category and stage of patient’s cervical cancer. The available data fit well three parameter Weibull distribution and the surviving probability of the patient significantly decreases. The survival times for non parametric and semi parametric approach each other while
there were higher mean survival times for parametric approach. Patient cancer stage significantly affected the survival of patients for each model than the other covariates. The results of this work showed non parametric and semi parametric methods were better performance to predict survival time of cervical cancer patients since median survival for both approach each other. Detection of cervical cancer at early stages and comprehensive treatment should be taken up to improve the overall survival of the patients as well as improve awareness in controlling cervical cancer.