Quantifying temporal variations in crustal materials associated with large earthquakes is critical for understanding the physical mechanisms (e.g., healing processes) of faults, therefore important for studies related to earthquake cycles. Repeating earthquakes that have nearly identical waveform recordings are a valuable tool for estimating temporal variations of the earth structures along the wave‐propagation paths of the repeating events. We have successfully applied a graphic processing unit (GPU)‐based template‐matching algorithm (cuNCC) to discover potential repeating earthquakes that occurred before and after the 2018 6.4 Hualien earthquake. After accurately relocating the repeating events and quantifying time shifts caused by location and origin‐time differences, two repeating event clusters (one doublet and one triplet) were identified for estimating the temporal variations related to the mainshock, based on our stringent qualifying criteria. A moving‐window cross‐correlation method was adopted to quantify the effects of the temporal changes on seismic waveform recordings. Our measurements show statistically significant delays after the mainshock for body waves, especially S waves, with wave‐propagation paths crossing the aftershock zone of the 2018 Hualien earthquake. In addition to body waves, measurements at some stations show significant delays for P‐ and S‐coda waves, suggesting that the scattering waves have also been affected by the temporal changes in material properties caused by the 2018 6.4 Hualien earthquake. We suggest the seismic community perform continuous monitoring of repeating earthquakes in the future for a more complete analysis to improve our understanding of the temporal material variations caused by the 2018 6.4 Hualien earthquake.