Abstract : GPS satellite and receiver differential L1/L2 biases induce errors into the GPS derived ionospheric measurements of the total electron content. If these measurements are to be used for ionospheric calibration of other systems or for deriving ionospheric measurements for these and other similar applications. A series of combined satellite plus receiver differential L1/L2 bias measurements has been made using phase smoothed pseudorange data from TI 4100 receivers at three sites operating over a five week period. These biases were estimated using a least squares estimation technique with a local second order polynomial ionospheric model. The analysis of these measurements shows that these biases can be estimated with an accuracy of 0.5 ns of differential delay, with improvements in the ionospheric model giving potential accuracies as low as 0.3 ns. The results indicate that these biases remain constant over a five week period. These measured biases will be compared with pre-launch calibration values and biases measured by other receivers. (rrh)