The study involved fire tests on twelve independent beamslab constructions of representative floor systems subjected to varying levels of end restraint. One assembly was designed and constructed noncomposite, two were designed and constructed composite, and nine were constructed pseudo-composite. The range of end restraint varied from zero to full end-fixity. The basic construction of all test assemblies consisted of a 4-in. concrete slab, 36 in. wide, cast over 22-gage steel deck. The concrete slab was supported by a 12 WF 27 ASTM A 36 steel beam, and the fire-exposed surface was protected with a spray-applied insulation. All materials were purchased on the open market, and the structural properties of the steel beams were as similar as could be obtained. The end support for the unrestrained test assemblies consisted of sliding and fixed-hinge bearings. The end connections for the restrained test assemblies consisted of either standard AISC B Series bolted clip angle connections or fully welded end plates. The typical B Series connections had the beam ends cut back 1/4 in. at each end from the back of the connection angles. One pair of B Series connections were made with the beam ends flush with the backs of the connections angles. Three assemblies were exposed to fire without end restraint and reached fire endurance end points at 90, 90, and 93 min, respectively. The major portion of the restrained pseudocomposite test assemblies reached fire endurance end points in the range of 111 to 113 min. Two test assemblies, which were more highly restrained, reached fire endurance end points at 107 and 106 min, respectively. On two asesmblies attempts were made to apply “optimal” levels of restraint designed to further extend the fire endurance performance. One assembly reached a fire endurance end point at 115 min, and the other assembly reached a fire endurance end point at 141 min.