Standardized, robust and time-efficient localization of the human secondary somatosensory cortex (S2) is a challenge in clinical blood oxygen level dependent (BOLD) functional magnetic resonance imaging (fMRI). A fully automated tactile stimulation was optimized in seven right-handed volunteers at 1.5 T for minimum scan time, high BOLD signals and robust localization of S2 by systematically varying the applied block-design. All volunteers had six different fMRI measurements of five stimulation-baseline-cycles (sbc) each with equal block duration that was changed between the measurements from 6 s to 30 s. Additional data sets of 4, 3 and 2 cycles were generated post hoc resulting in a total of 168 data sets that were evaluated individually for BOLD-signal intensity (dS%), correlation to the hemodynamic reference function (r) and Euclidean coordinates (x, y, z). Using different block-designs the S2 activation was highly variable regarding the localization rate (lr), the hemispheric symmetry and the BOLD-signals. The protocol with 3 cycles, a block duration (dp) of 15 s and a total scan time (dt) of 105 s most robustly localized S2 (contralateral: lr=71.4%, r=0.65, dS=1.01%; ipsilateral: lr=100%, r=0.6, dS=1.14%) whereas the most time-efficient protocol to localize SI (sbc=5, dp=6 s, dt=66 s) provided no robust localization of S2. Compared to other published fMRI protocols a scan time reduction up to 86% was achieved.