Dissolution-dynamic nuclear polarization (d-DNP) is a state-of-the-art technology that can dramatically enhance the detection sensitivity of nuclear magnetic resonance (NMR). DNP NMR has been applied to small molecules with stable isotopes and has been used to obtain metabolic and physiological information in vivo. However, the hyperpolarized state exponentially decays back to the thermal equilibrium state, depending on the spin-lattice relaxation time (T1). This signal decay has remained a major problem associated with this technology. Therefore, DNP NMR molecular probes useful for in vivo analysis have been limited to naturally occurring small molecules that inherently show long T1. While peptides are promising targets for DNP NMR studies, because of the limitation in T1, DNP NMR molecular probes applicable in vivo have been limited to amino acids or dipeptides. Herein we propose a 13C-labeling strategy to utilize the C-terminal [1-13C]Gly-d2 residue for realizing long T1 in peptides. Structure-based T1 relaxation analysis of amino acids and peptides revealed that (1) T1 does not decrease monotonically with increasing molecular weight and (2) T1 is not significantly affected by a side chain on the neighboring amino acid residue. These findings suggest that the C-terminal [1-13C]Gly-d2 residue affords sufficiently long T1 for biological uses, even in oligopeptides, and allowed us to develop 13C-b- casomorphin-5 (Tyr-Pro-Phe-Pro-[1-13C]Gly-d2, T1 = 24 ± 4 s at 3 T in H2O) and 13C-glutathione (g-Glu-Cys-[1-13C]Gly-d2, T1 = 58 ± 3 s at 3 T in H2O) as DNP NMR probes with long T1. We succeeded in in vivo detection of enzymatic conversions of these two probes. These results demonstrate the utility of our strategy and would contribute to further expansion of the substrate scope for DNP applications.