Measurements from three instrumented bridges in different parts of the United States were used to assess seasonal and diurnal behavior of concrete box-girder bridges. Effective bridge temperatures were used to evaluate seasonal thermal changes in bridges. Effective bridge temperature is defined as the temperature that governs longitudinal bridge movements. The effective bridge temperatures followed the same seasonal fluctuation as monthly average air temperature at the bridge site. Measured temperature differentials from top to bottom slabs for the three box-girder bridges were between+ 20ousF and-10ousF (+ 11.1 ousC and-5.6 ousC) regardless of geographic location. Temperature differentials were calculated by subtracting temperatures of the bottom slab from those of the top slab. Diurnal behavior of one bridge was continuously monitored for 24 hours during each of the four seasons. Longitudinal strains and temperatures were used to evaluate effects of nonlinear temperature gradient on bridge behavior. Internal thermal stresses at the instrumented bridge sections are reported.