The ability of matched-field tracking (MFT), a matched-field ambiguity surface summation technique, to track a source as it moves through the experimental region is demonstrated in this paper. The acoustic data are from two experiments: TESPEX 2, off the coast of Australia in 1994; and SWellEX-96, off the coast of California in 1996. Both were shallow water regions. Archival bathymetry and expendible bathy thermographs (XBT) taken during the experiment are used to construct the replicas corresponding to trial source positions in a horizontal plane. Matched-field processing is then used to match the data of a source track to the replicas for the region. MFT takes source motion into account by averaging values from these individual matched-field ambiguity functions corresponding to trial source tracks. This results in a new ambiguity function which depends on four parameters: the horizontal coordinates of the initial and final source position. Generally, the results are plotted as a function of the two horizontal coordinates of the final position. As increasing numbers of individual ambiguity functions are included, the movement of the source becomes apparent. The technique shows robustnesss to the uncertainties in replica computation (due to various mismatch problems) and periods when the signal to noise ratio is low.