Low-frequency repetitive transcranial magnetic stimulation (LF-rTMS) of the left temporo-parietal cortex (LTPC) has been proposed as a useful therapeutic method for auditory hallucinations (AHs). Stereotactic neuronavigation enables the magnetic coil to be targeted according to the individual parameters obtained from neuroimaging. Individualized rTMS neuronavigated according to 18-fluorodeoxyglucose positron emission tomography ((18)FDG PET) allows us to focus the coil explicitly on a given area with detected maxima of specific abnormalities, thus presuming a higher therapeutic effect of the method. The objective of this study is to test clinical efficacy of neuronavigated LF-rTMS administered according to the local maxima of (18)FDG PET uptake of LTPC and to compare it with treatment effects of standard and sham rTMS. In a double-blind, sham-controlled design, patients with AHs underwent a 10-day series of LF-rTMS using (1) (18)FDG PET-guided "neuronavigation," (2) "standard" anatomically guided positioning, and (3) sham coil. The effect of different rTMS conditions was assessed by the Auditory Hallucinations Rating Scale (AHRS) and the Positive and Negative Syndrome Scale (PANSS). Fifteen patients were randomized to a treatment sequence and ten of them completed all three treatment conditions. The intention-to-treat analysis of AHRS score change revealed superiority of the (18)FDG PET-guided rTMS over both the standard and the sham rTMS. The analyses of the PANSS scores failed to detect significant difference among the treatments. Our data showed acute efficacy of (18)FDG PET-guided rTMS in the treatment of AHs. Neuronavigated rTMS was found to be more effective than standard, anatomically guided rTMS.