Email plays vital role in the world of digital communication, still it has faced a huge challenge of harmful and malicious spam attacks. In spite a lot of measures and research contributions to guide against spam attacks, yet the spammers are adopting dynamic approaches to evade detection through content manipulation, text obfuscation and removal of hypertext tag which render the traditional methods such rule-based, blacklist ineffective. This research work embarks on evaluation of five widely used machine learning models- Naïve Baye’s (NB), Support Vector Machine (SVM), Decision Tree (DT), Multi-Layer Perceptron (MLP) and K-Nearest Neighbour (KNN) as well as their ensemble model which combined all these models in a single model for a better performance. The ensemble model performed better than each model.