This paper presents a novel method to placement of power quality monitors (PQMs) by using the multivariable regression (MVR) model, the Cp and Rp statistical indices. In this method initially, fault data in each bus is collected and then the correlation coefficient (CC) which show the relationship between buses during system disturbances are calculated to identify two or three buses with the highest CC values. These buses are considered as the most sensitive buses in the system. The identified bus voltages are then considered as independent variables in the developed MVR model to estimate the other bus voltages. After selecting two or three buses that have maximum and minimum frequency of the calculated CC, Cp and Rp, the appropriate number and placement of PQMs is then determined based on the lowest value of the Cp and a suitable value of the Rp. To validate the proposed PQM placement method, the IEEE 30 bus test system and the 69 bus test system are used.