Over the time, due to the variety of reasons such as economic problems and the fact realized by organizationsthat quality improvement may reduce the cost, Quality is considered as an important issue and the main factor ofevaluating the product or services by customers; so the quality is a key agent in order to accessing the betterCompetitive position. Products and processes generally have several related qualitative attributes because oftechnology development, promoting science and qualitative demands of customers. There has to be an appropriatemethod to control the attributes simultaneously especially when they are dependent. Multivariate quality controlmethods survey not only the effect of each variable but also the relation between them. Nowadays control charts areone of the best tools known to make the processes under control, detect appeared deviations and improvementpotentials. Specifying whether the data is in stable state -under control- or not, is dealt in the first phase of themultivariate process control procedure. In the second phase of the multivariate process control procedure, the rest ofprocess is being studied whether it is under control, using the achieved control limits from the first phase and futureobservations. So, the data, step change, unusual and outlier points existed in first phase may affect on control chartsof the second phase inconsistently, then, having found the outlier points of the first phase before the control limits tobe computed, is considered as an important issue. In this study, a robust estimator is issued applyinghierarchicalclustering technique that will detect the outlier points in multivariate control charts of the first phase in order to getthem removed. Then, the proposed method is evaluated by creating the variety of scenarios and compared with themost outstanding existed techniques. The evaluations represent that the proposed method performs better for themajority of simulated scenarios.