Fragments are generated from almost all types of warheads. The fragments are either primary fragments from warheads optimised for generating fragments or secondary fragments from warheads with other effects (pressure or penetration). Secondary fragments are generated by the protective material when projectiles or other types of penetrating warheads penetrates or perforates different protective materials. Warheads whose primary weapons effect is fragment are usually divided into three groups: natural fragmentation, controlled fragmentation or pre-fragmented. To be able to design protective measures against or estimate the lethality or vulnerability of the target from a fragmenting warhead it is necessary to characterise it. In this case characterisation is mainly to determine the mass distribution of the fragments, throw velocity and direction, shape and material density and strength. In the report, methods of characterisation and their respective advantages and disadvantages are compiled and discussed. For smaller warheads, up to about the size of a 155 mm artillery grenade, the existing saw dust pit is the best alternative. For larger warheads and tests in the field, board or hay/straw, in both cases together with witness plates, are the most suitable materials for soft-recovery of the fragments. Magnetic separation is the obvious choice for separating the fragments from the nonsolid soft-recovery materials. For non-magnetic metals eddy current-separators can use. If numerical simulations show good agreement with the experiments, parts of the experimental program aimed at determining the throw directions and velocities may be reduced and the experiments focused on establishing fragmentation. If the numerical simulations do not show good agreement with the experiments, flash X-ray combined with imaging plates or high speed cameras combined with flash X-ray and scintillation screens combined with computer aided image analysis should be used. Excess information from the writing of this report has been collected in an in-house report (Martin Nilsson, "Splitterkarakterisering v001", FOI, Stockholm, FOI-D--0329--SE, 2008).