The work presented involved testing of the compression strength of square specimens made from threelayer CLT. In one of the test set-ups a uniform compression over the complete square face of the specimen was used - similar to what is defined in the current European test standard for glued-laminated timber. In addition, several other test set-ups involving loading of only parts of the specimens’ square surface by line loads was investigated. The use of line loads aimed at investigating the effect of load distribution within the test specimen, and also to investigate possible boundary effects when the line load is applied close to the specimen edge. In practical design, the line load used in the tests, would correspond to the load transfer from a wall structure clamping a CLT-slab. During testing, apart from registering the load and deformation by conventional gauges, a contact-free deformation measurement system was used. The results show that the compression strength – defined according to Eurocode 5 as the load at which a 1% permanent set is obtained – is indeed dependent on the relative size of the load application area, its orientation relative to the surface grain direction and its distance to the edge of the specimen. Other conclusions are that test methods, test evaluation methods as well as design criteria for compression perpendicular to grain in CLT should be revisited.