Several functional MR imaging studies evaluating the lateralisation of linguistic functions in patients who underwent Wada testing have been reported. There is extensive variance in the Laterality index (LI) calculation across the studies, and the optimal calculation method remains unclear. We attempted to calculate the LI in different ways in the same subjects, in order to find the LI calculation method with the highest correlation to the Wada test. Fifteen patients (10 females, 5 males) suffering from medically intractable temporal lobe epilepsy (TLE) (12 left, 3 right) were admitted for the study. The patients underwent a standardized bilateral intracarotid short-acting barbiturate test. Language testing included spontaneous speech, oral comprehension, reading, object and picture naming, and repetition. All the tasks were scored separately in order to increase the possibility of correlation between Wada and LI. A silent phonemic verbal fluency task (VFT) was used as a language paradigm for functional measurement. Regions of interest (ROIs), with a known association with language function (Broca's area, the lateral prefrontal cortex, etc.), were defined. First, the LIs were calculated from the ROIs using a previously reported method (simple suprathreshold count). Next, we used several new methods of LI calculation (t-weighting of voxels, methods independent of the choice of the statistical threshold, etc.) The most significant correlation with Wada was proven in the LIs that were evaluated from Broca's area (up to R = 0.94, P = 1 x 10(-7)). However, the new LI calculation methods used in the present study did not produce a statistically significant benefit in comparison to previously reported methods.