The aim of this paper is to analyze the performance of an intrusion detection model using single- and ensemble-based classifiers. Several tree-based single classifiers were analyzed. The ensemble of tree-based classifiers was also analyzed to differentiate the superiority in their performance. Different proportions of the benchmark KDD dataset are utilized for observing the performance of the model. Classification based on the accuracy, model building time, and kappa statistic is evaluated as the performance measures in this paper. The base and ensemble classifiers resulted in better accuracy are observed in the experiments and only Naive Bayes and random tree resulted in minimum model building time. Most of the classifiers produced better results for kappa statistic. The highest statistic is computed for ADA classifier, and lowest error is computed for the random forest ensemble. © 2020, Springer Nature Singapore Pte Ltd.