For life cycle analysis, it is essential to know the structure capacity at a given point in time. Determining the strength from equations that do not consider the size effect can lead to insufficient maintenance plans, higher costs and potentially unsafe structures. With this aim, the 'Type 2' size effect, valid for structures that develop large cracks prior to reaching the maximum load, and 'Type 1' size effects, for structures with no pre-existing crack, have each been widely investigated and are quite different. In practice, however, a pre-existing crack at maximum load can be neither large nor negligible compared to the structure dimensions. Therefore, the Type 1-Type 2 transition is important. To determine it, and clarify the role of crack length, a large-scale experimental program, consisting of more than 120 geometrically similar three-point bend beams of 4 sizes, with the size range of 1:12.5, as well as 24 cylinders and 12 beams for ASTM flexural strength tests, has been undertaken. At the time of writing, only partial results are available. A detailed journal article is in preparation.