This paper analyzes some issues related to the metrological characterization of instruments for power quality monitoring, in order to move a step toward the definition of a full performance verification protocol. This protocol should include not only the test situations reported in corresponding standards, but also an array of different voltage and current fluctuations -linked to common power quality issues -that may be encountered in actual power systems. This performance analysis is particularly interesting because it can be found that different power quality instruments, fully meeting characteristics prescribed in standard, still may significantly disagree in some actual measurements. This is due for the existence of unresolved technical and theoretical issues related to the metrological characterization through the classical approaches of these instruments. The aim of the discussion carried out in this paper is also the specification of requirements of a test system devoted to calibration and verification of a power quality monitor, such as it is done in type testing. After a preliminary discussion about technical and theoretical issues related to the performance analysis of these instruments, a proposal of test protocol definition, with reference to some power quality phenomena and based on the design of experiment, is presented.