High‐performance fibre reinforced concrete represents a promising alternative to conventional reinforced concrete for the construction of prestressed precast girders for road bridges. Specifically, all shear reinforcement is replaced by the use of steel fibres, which leads to important savings from the material and economical perspectives. Therefore, reliable shear formulations are required for these specific cases. In an effort to evaluate the accuracy of code expressions, this paper presents a database collecting 80 experimental shear tests on prestressed fibre reinforced concrete beams without stirrups. After illustrating its main characteristics, this database is used to verify the conservativeness and accuracy of shear formulation according to several codes and recommendations: i) Model Code 2010, ii) draft version of Model Code 2020, iii) draft version of the new Eurocode 2, and iv) Spanish National Code. The evaluation, carried out through the calculation of the model safety factor, indicated that the application of all shear formulations is largely conservative, leading to less unsafe tests than the desired 5% fractile. The relation between model safety factors and mechanical parameters is then investigated. Finally, the results obtained for the developed database are compared with those achieved from two larger databases of beams with similar features but without prestress. The comparison clearly proves the greater conservatism of code expressions when beams are prestressed.