When structures are tested against earthquake loading, either within laboratories or by computer modelling, observed strong-motion records are commonly used as input. However, the method to select records is often not stated and it is thought that records are often chosen simply because they are well-known. Until the early 1990s, strong-motion data was relatively difficult to obtain and process and, in addition, the amount of, particularly near-source, strong-motion was limited. Therefore using standard records was often the only option. However, since then a number of large well-recorded earthquakes have occurred and also there are many new strong-motion data sources available. As a snapshot of the state-of-the-practice for record selection for structural testing in Europe the proceedings of the last European Conference on Earthquake Engineering (12ECEE) was used. Using those articles concerned with structural testing where strong-motion records are used, a summary of the most popular records chosen and, if known, why they were selected, is given. It is found that a small subset of 13 well-known records is often used, irrespective of other considerations. The mode of the number of input accelerograms used per study is three. Strong-motion records within structural testing could be used simply to validate that the shake table or the computer model functions and to show general aspects of the structures' response. For such applications the choice of the strong-motion records is not important and the current method of using well-known records could be maintained. However, more often structural testing is employed in order to understand the characteristics of the modelled structure and to estimate its vulnerability to earthquake shaking. For this purpose a more scientific method of choosing input time-histories is required. The procedure suggested here recognises the observed variety with respect to amplitude, frequency content, duration and shape of earthquake ground motions in order to better understand the characteristics of the structure.