The current groundwater monitoring network established and operated by the Water Resources Agency has successfully provided valuable information on the main aquifers in Taiwan. However, the cost of maintaining extensive monitoring of both the levels and quality of groundwater is very high. As a result, developing a cost-effective program for monitoring the quality of groundwater, which involves sampling from only a fraction of the existing monitoring wells, is important. Such a monitoring program can yield sufficient information to determine the quality of ambient water quality in the main aquifers at reduced cost. This work implements the multivariate geostatistical method to select a well network for monitoring groundwater quality indexes under considering multi-variables of multi-scales geostatistical structures. The results of this study are the rank of original groundwater level monitoring wells in the Pingtung plain as groundwater quality index wells. The results of this study indicate that the multivariate geostatistical method is suitable for locating groundwater quality index wells that provide sufficient information at reduced cost. The application of the proposed approach can be extended to other regions.