Outliers can significantly affect data mining performance. Outlier mining is an important issue in knowledge discovery and data mining and has attracted increasing interests in recent years. Class outlier is promising research direction. Few researches have been done in this direction. The paper theme has two main goals: the first one is to show the significance of Class Outlier Mining by discussing a comparative study between a Class Outlier detection method called Class Outlier Distance Based (CODB) and a conventional Outlier detection method. The second goal is to introduce Enhanced Class Outlier Distance Based (ECODB) algorithm which is enhancement of CODB algorithm. ECODB reduces CODB parameters using a heuristic approach. The experimental results show that CODB can detect Class Outliers that cannot be detected using conventional Outlier detection methods. The experiments also show that ECODB works efficiently as CODB.