Over the past three decades, wind turbine towers have experienced a rapid development. Current constructions like the so-called hybrid towers have reached heights up to 165 m. These hybrid towers usually consist of steel sections in the upper part and a segmented prestressed concrete tower in the lower part. The horizontal joints in the concrete towers are usually designed as dry joints without shear notches or force bearing connections. The external pre-stress tendons are constructed inside the tower. The shear forces due to transverse forces and torsional moments are only transmitted by friction of the joints. In particular, the torsional capacity of this construction represents a design relevant parameter. First investigations have shown that the classical torsion theories do not provide satisfactory results for this type of construction. The discrepancies of the theoretical methods and numerical simulations were investigated and now presented. A large-scale test of a segmental reinforced concrete tower was carried out to a scale ratio of ~1:10 in comparison with a real in-site tower. The test results are presented in relation to current theories and numerical simulations. These results show that the current theories do not correctly estimate the joint torsional bearing capacity. More specifically, the current theories deliver unsafe results when the joint is closed and unused potential when the joint gapes. With regard to the numerical calculations, the load-bearing capacities could not be conclusively confirmed either. The test-determined load-bearing capacity here is lower than simulated results.