Email spam is one of the biggest challenges when using the Internet today. It causes a lot of troubles to users and does indirect damages to the economy. Machine learning is a keyapproach for spam filtering. Artificial Immune System (AIS) is a diverse research area that combines the disciplines of immunology and computation. Negative selection mechanism is one of the most studied models of biology immune system for anomaly detection. In this paper, Negative Selection Algorithms (NSA), a computational imitation of negative selection, ismodeledfor spam filtering. The experimental results on popular TREC'07 spam corpus show that the approach is an effective solution to the problem on both time complexities and classification performance.