The motor threshold (MT) is a fundamental parameter for evaluating cortical excitability in transcranial magnetic stimulation (TMS) despite remarkable variation, both within, and between subjects. We intended to test whether the variation could be reduced by targeting the stimulation on-line and modeling the TMS-induced electric field on individual MR images. Navigated TMS was used to map the primary motor cortex for the representation area of the thenar muscles (abductor pollicis brevis) and to determine the MT. Thirteen healthy subjects participated in the study. To determine the between-subject variation, the MTs of nine subjects were measured with two different stimulators (comparison study). To study the individual variation, the MT measurement was repeated 20 times in four subjects always using the same stimulator (longitudinal study). In the comparison study, the MTs differed significantly between the two stimulators over all subjects (p<0.001), whereas the electric field strengths did not exhibit significant difference between the stimulators. Both, the MTs, and the electric field strengths showed similar variations, which were greater between subjects (comparison study) than within subjects (longitudinal study). In the comparison study, the distance between the locations of the two different coils on the scalp was significantly greater than the distance between the induced electric field maxima in the brain (p<0.001). We conclude that on-line navigation can be used to reduce the variation caused by different stimulator types and individual subject anatomy. In addition, cortical excitability can be evaluated by using computed electric field strength as well as stimulator-dependent MT.