Transcranial magnetic stimulation (TMS) can be used for non-invasive assessment of cortical physiology and descending motor pathways. However, the focus/exact site of cortical activation is considerably widespread in traditional TMS. When combined with MRI-based navigation, it allows specific anatomical areas of the cortex to be stimulated. The peripheral muscle responses to TMS are commonly measured as motor evoked potentials (MEPs). We compared the accuracy of cortical mapping, as well as the congruity of the motor thresholds (MT) and MEPs between navigated and non-navigated TMS procedures. Eight volunteers were studied in two sessions. In each session both hemispheres were stimulated with and without navigation. Non-navigated TMS: Both hemispheres were mapped without navigation to find the representation area of the thenar muscles based on induced MEP amplitudes. MT was then determined at the optimum coil location. Navigated TMS: Individual MR-images were used for the on-line navigation procedure. The cortical representation area of the thenar musculature was mapped at the "hand knob". The optimum stimulus target was used for MT determination. The order of these two procedures was randomized. Following the MT determination, MEPs were recorded from 20 consecutive stimuli. The MTs were similar from session-to-session with no inter-hemispheric differences, and with and without navigation. The stimulus location was more spatially discrete in navigated TMS producing more stable MEPs with significantly higher amplitudes and shorter latencies. In summary, MEPs exhibit significant differences depending on whether navigation is used. However, the MTs are not significantly dependent on the discrete stimulation site.