Recent interest in Darrieus wind turbines has led to a need for proper performance prediction models. Of all the existing models, multiple streamtubes model is reasonably simple and relatively precise. However, neglecting dependency of Reynolds number (Re) leads to the failure of prediction at high tip-speed ratios (TSR). In this proposed paper, an algorithm LDWT, which applied local blade Re instead of one representative Re, was developed and researched. In the calculation, 2-dimensional experimental data for aerofoil characteristics were applied because their 3-dimensional data for different Re could not be accessible. This study shows that results for high TSR from LDWT match test data better than previous research instead of underestimation in other regions. In the region of low TSR, it is shown that stall delay model is needed to match well with test data. Therefore, it is expected to be useful in the proper design and optimisation of rotors at high tip-speed ratios when their test data is not available or incomplete. (C) 2012 Elsevier Ltd. All rights reserved.