: Process capability analysis is one the most important subjects in quality topics. In fact process capability study has a special role in any quality improvement problem. Process capability indices using thespecification limitsare the most important tools used in process capability studies. These indices evaluate the accuracy, precision and the performance of manufacturing process. During the last decades properties of univariate processes have been investigated extensively, but are comparatively neglected for multivariate process where multiple dependent characteristics such as weight, length etc. are involved in quality measurement. Nowadays with highly sophisticated advanced measuring instruments, Access to higher volumes of qualitative and quantitative data are provided and therefore these indices have wider application in industries. On the other hand the quality of data on the process characteristics relies very much on the gauge measurement capability, therefore the capability indices of the process by ignoring the measurement errors would not be reliable. According to the role of the value of these indices in suppliers selection and determining the acceptance sampling programs, ignoring the gauge capability may lead to significant deviation from the accurate value of these parameters and producer will suffer if they ignore this fact. It is also shown that one can obtain a better estimation by considering the gauge measurement capability. Multivariate process capability indices unlike the univariate process capability indices are not unique and are calculated based on different definitions. So there is a bewilderment in selecting appropriate index for evaluating multivariate process. In this study six different multivariate indices are investigated and a procedure is proposed to calculate multivariate indices considering gauge capability. With the purpose of determining the robust index, the sensitivity of each index were investigated against the gauge measurement error and finally the and indices were introduced as the robust indices respectively for the indices defined based on ratio of tolerance region to the process region and the indices defined based on principle component analysis.
