We aimed to evaluate the biodistribution and specificity of Ga-DOTA-TAT and RHO-TAT using MGC-803 and HT-29 tumor cells as well as tumor-xenografted nude mice and to demonstrate its application in positron emission tomography (PET) imaging. The in vitro evaluation of Ga-DOTA-TAT was assessed in MGC-803 and HT-29 cell lines, and the in vivo evaluation of Ga-DOTA-TAT was also performed in mice bearing MGC-803 or HT-29 tumors, respectively. Fluorescence microscopy was also employed to evaluate the specificity of RHO-TAT in vitro in MGC-803 and HT-29 cells as well as ex vivo in tumor slices of the corresponding tumor models. The in vivo imaging differences between Ga-DOTA-TAT and F-FDG in MGC-803 and HT-29 tumors were also studied. The biodistribution and micro-PET results demonstrated significant uptake of Ga-DOTA-TAT in non-FDG-avid MGC-803 tumors, whereas there was negligible uptake in FDG-avid HT-29 tumors. RHO-TAT showed superior fluorescence microscopy imaging effects in MGC-803 cells and tumor slices but not in HT-29 cells and tumor slices, which were consistent with the in vivo results. Ga-DOTA-TAT combined with F-FDG can be applied noninvasively in cancers with PET imaging for potential patient selection and stratification. We demonstrated a higher binding of Ga-DOTA-TAT and RHO-TAT to MGC-803 cells as well as to non-FDG-avid MGC-803 xenografted tumors and a lower binding to HT-29 cells and FDG-avid xenografted tumors. These results suggest that TAT has the potential to be a ligand for targeting certain tumors.