This study investigates the accuracy of spatially resolved skin friction velocity ut measurements using a large number of Irwin sensors (Irwin, 1981). A fast and simple procedure to calibrate the sensors against a two-component hot-film anemometer is introduced which leads to a universal calibration function that can then be used for all sensors. Insignificantly lower accuracies are obtained using the universal calibration function instead of individual calibration functions for each sensor. The overall accuracy of the skin friction velocity measurements averages about +/- 5% for u(tau) > 0.13 m s(-1) and decreases strongly to +/- 20% at lower u(tau). The main sources of errors are as follows: (i) deviations from flush mounting of the sensors with the wall surface, (ii) inhomogeneous boundary layers produced in wind tunnels, (iii) pressure transducer inaccuracies, (iv) Irwin sensor geometry variations and (v) Irwin sensor calibration errors. An excellent linear relationship between u(tau) measured with the Irwin sensors and the free stream velocity as well as the good agreement of Irwin sensor with independent hot-film measurements validate the reliability of the measurement setup. An exemplary measurement of the skin friction velocity distribution around a single wall-mounted rectangular block is presented, discussed and compared to literature and can serve as a test case for simulations. (C) 2012 Elsevier Ltd. All rights reserved.