In this paper, we present a new model for the measurement of yield curve relationships that is derived from interest rate theory, utilizes an objective procedure, and provides measures of the accuracy of the results obtained. In empirical tests of the model, the structure postulated is found to consistently provide a high level of explained variation in observed market yields on U.S. Treasury bonds. In a comparison with a yield curve model previously offered by Cohen et al., the present model is superior in terms of both goodness of fit and other associated statistical criteria. Clear evidence exists that the impact of coupon level upon yield is statistically significant, consistently positive in direction, substantial in magnitude, and variable over time. These results indicate that correction for coupon differences in the calculation of forward rates for use in empirical tests of interest rate theory is necessary in order to obtain reliable results. Finally, the yield curve model is used to calculate estimates of the risk-free pure discount rate.