In research, and in analysis for design, the global and local behaviours of bridge deck slabs are normally considered separately. Research has shown that bridge deck slabs are able to resist very much greater local wheel loads than is implied by conventional design methods and, as a result, new empirical design methods have been proposed. Bridges with the very lightly reinforced slabs designed to these rules have
not previously been tested under full NB load. Non-linear analysis suggests that global transverse moments could significantly reduce the local strength of bridge deck slabs, and tests on two half-scale model bridge decks confirm this. Thus both the tests and the analysis show that global and local behaviour are not independent. Several other aspects of the behaviour of the bridges which conflict with the suggestions of previous research are also predicted by the analysis and confirmed by the tests. Despite this, both the analysis and the test results show that the behaviour of bridges with deck slab reinforcement designed to the empirical design rules proposed for Britain will be satisfactory. It even suggests that some of the requirements previously imposed for the rules, notably the requirement for support diaphragms, are unnecessary.