This paper proposed a new multivariate correlation analysis (MCA) based detection system to protect online services against Denial of Service (DoS) attacks. It is powered by the triangle-area based MCA technique and the anomaly-based detection technique. The detection process consists of three major steps: Basic Feature generation for network traffic records, multivariate correlation analysis, and anomaly based detection mechanism. Evaluation has been conducted using KDD Cup 99 data set to verify the effectiveness and performance of the proposed DoS attack detection system