This research aims to find the multivariate control chart which is suitable for production process, determine the current capability process, and also find the dominant variable which is correspondent to the improvement of drinking water quality. This research use a secondary data collected from January 2016 until February 2017. There were four group of variable used in this research. The first group was consist of color, turbidity, maximum containing level of copper and manganese. The second group consist of level of total dissolved solid, hardness, calcium and magnesium. The third level was consist of pH. value and alkaline; and fourth group consist of nitrate and nitrite. The result of data processing showed us that Multivariate Exponentially Weighted Moving Average or MEWMA control chart can detect the out of control data in the each group of variable faster than T2 Hotelling control chart. Then, the capability process index for the first group of variable is 2.238, the capability process index for second group of variable is 4.208, the capability process index for third group of variable is 1.438, and capability process index for fourth group of variable is 1.346. All group of variable have capability process index more than 1. This condition indicated that process goes well in all group of variable although there were some data is out of control in each group variable.