The role of renal sympathetic nerve (RSN) in hypertension should be better understood. We aimed to three-dimensionally reconstruct the renal nerves, and explore its anatomical and histochemical characteristics in hypertensive canine model and patients. Renal arteries with surrounding tissue were collected from canines and cadavers with or without hypertension. Serial renal artery hematoxylin-eosin sections were used for three-dimensional reconstruction, and morphological parameters were collected and analyzed. In hypertensive canines, the mean renal nerve number was 26.71 ± 5.68 versus 19.84 ± 5.68 in controls (P = 0.02), and the middle renal nerve volume was 5.31 ± 2.13 versus 2.60 ± 1.00 μl in controls (P = 0.01). Renal tissue norepinephrine concentrations, tyrosine hydroxylase and substance P immunoreactivity in RSN, and growth-associated protein 43 immunoreactivity in renal ganglion were significantly increased in hypertensive canines. In humans, the renal nerve was evenly distributed along the renal artery in a network pattern. The renal ganglion volume was 72.75 ± 33.43 in hypertensive patients versus 37.04 ± 23.95 μl in controls (P = 0.029) and the mean neuronal size in renal ganglion was 1187.3 ± 219.9 μm in patients versus 714.8 ± 142.7 μm in controls (P = 0.002). Tyrosine hydroxylase immunoreactivity in the RSN was 0.153 ± 0.014 in patients versus 0.104 ± 0.019 in controls (P = 0.013). Growth-associated protein 43 immunoreactivity in the renal ganglion was 86 612.8 ± 14 642.0 in patients versus 33 469.8 ± 15 666.8 μm/mm in controls (P < 0.001). Our study suggests that RSN and renal ganglion histological remodeling occurs in individuals with hypertension and the distal segment or branches of renal artery might be a promising therapeutic target for RSN modulation therapy.