Cervical cancer, a threat to female existence is one of major cancer affecting women in the developing countries of the world. Several factors are responsible which humans didn™t take cognizance of. These factors are numerous and can at times be difficult to explain using linear regression because it can™t handle many dummy variables that are not necessary to create qualitative predictors. This study uses decision trees to classify and identify the major risk factors causing cervical cancer in women depending on their age since it closely mirrors human decision making than the classical regression approach. A regression tree was constructed from the training data using recursive binary splitting. There was a minimum number of observations required for each terminal node before it stopped. Then cost complexity pruning to the large tree in order to obtain a sequence of best sub trees was applied. By using decision trees as building blocks, we can construct more powerful predictions for decision trees, bagging, random forests, and boosting. 858 cervical cancer patients were observed using 34 risk factor attributes from University Hospital of Caracas, Venezuela. Using classification trees, 14.22% of errors are produced during training. Based on the test data set, 91.5% of the predictions are correct. Based on the data set’s pruned data, 91.75% of the observations can be classified correctly. Test predictions generated by this model are within 67 years of the true median age of patients, based on regression trees. Bagging and Random forest show improvement on the regression trees by setting a reduced mean square error. There are four most significant variables among all trees examined by the random forest, including age at first sexual intercourse, number of pregnancies, number of sexual partners, and hormonal contraceptives. The same goes for boosting, as a result of the relative influence statistics.