BBRI is carrying out a research program in order to evaluate the recommended values proposed by Eurocode 6, in the context of the establishment of its National Annex. This paper concerns the compressive strength. After a first series of tests on masonry made of clay units (groups 2 and 3), a general observation was that the formulations for combinations with general purpose mortars were unsafe, while those with thin layer mortar were (too) safe. In order to understand better the reasons of these differences, the influence of the shape factor of the unit was further investigated experimentally. The research program was based on compressive strength tests on cut and uncut units and also on masonry walls made of the same cut and uncut units. The full range of tests was been performed for one type of clay unit and one type of concrete unit in combination with several mortars. The conclusions are that the shape factors are not realistic for the types of units investigated in this research. It's suggested to investigate further the dependence of the shape factor on the grouping of masonry unit according to Eurocode 6. Moreover, for the studied geometries of units, the experimental loss of compressive strength when using half units instead of whole units in walls tested according to EN 1052-1, instead of being equal to 0%, fluctuates from 6 to more than 50% depending on the combined materials (unit and mortar). The fact that the use of different shapes of a same material in masonry influences the compressive strength is thus not sufficiently taken into account in Eurocode 6. Because of the number of parameters playing a role in the EC6 formulations, these research activities constitute an interesting contribution to other complementary studies.