Before 2007, the Minnesota Department of Transportation (MnDOT) found that many precast, prestressed concrete bridge girders were arriving at the construction site with cambers much lower than predicted. It was thought that the method for estimating the camber at the time of erection was the problem. The method being used at the time was the �PCI multiplier method,� which specified that the upward deflection due to initial prestress and downward deflection due to self-weight be multiplied by 1.80 and 1.85, respectively. In 2007, MnDOT began using a single multiplier of 1.5, but observed erection cambers were still lower than predicted. In a study conducted by O�Neill et al. (2012), it was found that, for 1067 girders produced between 2006 and 2010, the average camber at release on average was only 74% of the predicted design camber. The main factor contributing to the lower than predicted cambers was the underestimation of the concrete elastic modulus at release. The underestimation of the elastic modulus resulted from two factors. First, the relation used to predict the modulus was the ACI 363 equation associated with high strength concrete (i.e., f�c = 6,000 psi) rather than the Pauw (1960) equation. Second, the concrete compressive strength at release used in the equation was underestimated by approximately 15% on average. These factors accounted for the majority of the difference in the measured to predicted cambers. The remainder of the difference was attributed to possible prestress losses due to thermal effects during fabrication. Because the girders are cast on a fixed bed, changes in temperature during fabrication cause associated changes in stress in the strands. In this study, it was assumed that the temperature changes that occur prior to steel/concrete bond result in unrecoverable changes in prestress force that become locked in the girders. This report summarizes an