Numerous studies have shown that basic Tat peptide (48-57) internalized non-specifically in cells and localized in the nucleus. However, localization of imaging agents in cellular nucleus is not desirable because of the potential mutagenesis. When conjugated to the peptides that undergo receptor-mediated endocytosis, Tat peptide could target specific cells or pathologic tissue. We tested this hypothesis by incorporating a somatostatin receptor-avid peptide (octreotate, Oct) and two different fluorescent dyes, Cypate 2 (Cy2) and fluorescein 5'-carboxlic acid (5-FAM), into the Tat-peptide sequence. In addition to the Cy2 or 5-FAM-labeled Oct conjugated to Tat peptide (Tat) to produce Tat-Oct-Cypate2 or Tat-Oct-5-FAM, we also labeled the Tat the Tat peptide with these dyes (Tat-Cy2 and Tat-5-FAM) to serve as positive control. A somatostatin receptor-positive pancreatic tumor cell line, AR42J, was used to assess cell internalization. The results show that Tat-5-FAM and Tat-Cypate2 localized in both nucleus and cytoplasm of the cells. In contrast to Tat-Oct-Cypate2, which localized in both the cytoplasm and nucleus, Tat-Oct-5-FAM internalized in the cytoplasm but not in the nucleus of AR42J cells. The internalizations were inhibited by adding non-labeled corresponding peptides, suggesting that the endocytoses of each group of labeled and the corresponding unlabeled compounds occurred through a common pathway. Thus, fluorescent probes and endocytosis complex between octreotate and somatostatin receptors in cytoplasm could control nuclear internalization of Tat peptides.