This study evaluated two training procedures that might be used to increase native English (NE) and native Chinese (NC) listeners' ability to discriminate the mid vs. low tone contrast in Thai under two different inter-stimulus-interval (ISI) conditions (i.e., 500-ms and 1500-ms). Participants were assigned to receive training using either a two-alternative forced-choice identification (ID) procedure or a categorial same/different discrimination (SD) procedure. Preliminary results from 14 NE and 14 NC listeners showed significant increases in percent correct identification in both ISI conditions for NE listeners who received a two-alternative forced choice training. However, performance of the NE listeners who were administered a same/different categorial discrimination training procedure did not improve significantly after training in either ISI condition. On the contrary, NC listeners seemed to have benefited equally from the two training procedures. That is, a comparable and significant improvement in Thai tone discrimination were observed among NC participants in both the ID and the SD groups. Additionally, no effect of ISI was observed in either group of listeners either before or after training. These results suggest that the identification training procedure is superior to the same/different discrimination method in promoting phonological representation of Thai tones among English listeners, but not among Chinese listeners. However, more data, which is being collected, is needed to confirm these results. If upheld, these results suggest that prior exposure to tones may facilitate the formation of foreign tones, thus causing the differential benefit of the two training procedures to be minimized.