As GABA editing is achieved by eliminating the overwhelming Cre peak using subtraction, discrepancies in Cre amplitude due to physiological noise result in variations in edited GABA. It can be observed the flip of the outer peaks of GABA triplet is in a different dimension from the Cre. It is thus possible to eliminate Cre and its associated physiological noise using principle component analysis (PCA). Simulation and phantom studies showed the physiological noise associated with Cre was eliminated using PCA editing without GABA signal loss. In vivo study demonstrated more consistent GABA measurement using PCA editing than regular editing.