Measurements of loads and bending moments on concrete crossties for several days of revenue traffic were used to develop a statistical description of track loads for tangent and curved tracks that have variable tie spacing. The measured data show large tie-to-tie variations in loads and a load-dependent tie support condition. Many ties were center-bound for loads from light or empty cars, but the tie support became more uniform for heavy wheel loads. Maximum tie bending moments measured on curved track were considerably higher than those on tangent track because of the increase in vertical and lateral loads on the high rail when trains exceed the balance speed of the curve. Tie bending moments measured in this program were considerably lower than the current static flexural strength requirements for a probabilistic prediction of maximum loads for a 50-year life. These and data from other concrete-tie test installations indicate a need to identify the failure mechanism for concrete ties so that statistical load descriptions can be used for future design and testing. Low-probability maximum loads will be very important if failures result from infrequent loads that exceed the static strength. However, the higher probability mean cyclic loads will be the more important factor if fatigue is identified as the governing failure mechanism.