All available data on unintended composite action in beam-andslab bridges are reviewed and the factors that may influence the existence of unintended composite action in noncompositely built beam-and-slab bridges are investigated. Test reports summarized in this paper have shown that the existence of a natural or a chemical bond is the single most important factor in determining whether a noncompositely built beam-and-slab system can be counted on to act compositely. There is considerable evidence that indicates the presence of composite action in bridges in which no provisions were made for such action. This composite action may reduce the stress under a given load by a significant amount. However, the uncertainty surrounding the presence of composite behavior and the difficulty associated with verifying the existence of composite behavior make the assumption of composite behavior in a bridge designed noncompositely a questionable one.

An opinion that appears to be widely held by structural engineers involved in the design or the rating of bridges is that composite action will exist in a steel beam-concrete slab bridge whether such action was provided for in the design of the bridge or not. Bridge 4 in the Tennessee bridge tests (I) was a steel girder bridge that was designed to act noncompositely but that, in fact, acted compositely up to the load at which yield in the steel girders would have occurred in the noncomposite bridge. This rather widely publicized result has contributed to the widely held opinion just described. Unfortunately, consideration of all available data does not permit the drawing of a general conclusion in this regard.