The resistance to delamination of cross-laminated timber (CLT) made from spruce has been investigated. The main goal was to examine if delamination results are influenced by the shape of test specimens as both square and round specimens may be evaluated according to the CLT standard EN 16351. Furthermore, the influence of the production parameters number of layers, layer thickness and bonding pressure on delamination behavior were investigated. The results show that specimen shape and number of layers affect the resistance to delamination. Square specimens and a high number of layers show significantly higher delamination than round specimens and specimens with low number of layers. Layer thickness and bonding pressure, however, have no influence. As the number of specimens which exceeded delamination or wood failure requirements given in the CLT standard was very high it is questionable, if the applied delamination procedure is suitable for evaluating the bond integrity of CLT. As a result, it is recommended to eliminate the option to use square or round specimens in delamination tests and to revise the CLT standard accordingly. Furthermore, we suggest either to adjust the delamination test parameters or to develop a new test method to be able to assess bond durability more realistically.