The sensitivity to electrode misplacement of the 12-lead ECG, synthesized from differential leads, was analyzed by simulating the electrode misplacement, with a help from interpolated body surface potential maps. The electrodes forming optimal differential leads were independently moved only on the lines defined by gradient vectors, calculated on the integral body surface potential map, obtained from juxtaposed average beats for each of the 47 multichannel ECGs used in the study. The 12-lead ECGs were synthesizes from the simulated moved differential leads and compared to the ECGs synthesized from the nominal electrodes positions, by using correlation coefficients and RMSD. The results show that synthesized 12-lead ECGs are less accurate if electrodes with bigger associated gradients are misplaced, but also suggest that small displacements of up to 1 cm do not have a big influence on the synthesis quality. From the perspective of ECG synthesis, the presented methodology can be used to define an additional criterion for optimal triplets of differential leads selection, which will ensure that the differential lead triplet are optimal not only in terms of the quality of the synthesized leads, but also in terms of being minimally sensitive to electrode misplacement.