A 78-year-old man was admitted to a local hospital because of leg swelling and polyuria (both lasting for approximately 2 months). He had not received any medication for his complaint. He had no known medical history of hypertension, diabetes mellitus. Physical examination showed the patient to be well nourished but he craved for water. Retroperitoneal fibrosis was detected on computed tomography (CT). The retroperitoneal fibrotic tissue was compressing both distal ureters, and bilateral hydronephrosis was identified on CT. To resolve the hydronephrosis, a double J stent was inserted in the right-side ureter; however, insertion into the left-side ureter failed due to atrophy. To evaluate the cause of the retroperitoneal fibrosis, we decided to perform a laparoscopic surgical biopsy. Because the retroperitoneal fibrotic tissues were too stiff for biopsy, we biopsied the para-aortic lymph node, just adjacent to the fibrotic tissue. The biopsy test results showed the classical characteristics of Castleman’s disease (hyaline-vascular type, negative for human herpesvirus 8). Before the initiation of radiation treatment, he still complained of frequent urination (20 times/day), excessive thirst (visual analogue scale 8) [], and his urine output was approximately 5~6 L/day. His baseline biochemical levels were as follows: blood urea nitrogen, 19.6 mg/dL; creatinine, 1.4 mg/dL; sodium, 149 mEq/L; potassium, 4.8 mEq/L; chloride, 118 mEq/L; serum osmole, 311 mOsm/kg; and random glucose, 131 mg/dL. On urine analysis, the specific gravity was under 1.005, representing diluted urine, and the urine osmolality was 148 mOsm/kg. No protein, glucose, or red blood cells were seen on a urine analysis. A basal plasma AVP level was 5.24 pg/ml, which was above the normal range (0~4.7 pg/ml). An antinuclear antibody test was performed to further evaluate the retroperitoneal fibrosis. The antinuclear antibody test was positive, with a homogeneous pattern, but the specific tests for extractable nuclear antigen antibodies and double-strand deoxyribonucleic acid were negative. A water deprivation test was not performed, as his serum osmolality (311 mOsm/kg), and serum sodium (149 mEq/L) were above the threshold for maximal AVP secretion (serum osmolality, 300 mOsm/kg; serum sodium, 145 mEq/L). Thus, we performed a vasopressin challenge test (Additional file: Figure S1). After the vasopressin injection, his urine osmolality increased to 206 mOsm/kg, which was approximately 39% greater than that at baseline (before the vasopressin injection: 148 mOsm/kg). Thus, the urine was not sufficiently concentrated to the expected range, indicating partial nephrogenic DI. Partial nephrogenic DI can be diagnosed as a small elevation (up to 45%) in urine osmolality after a vasopressin injection, with the urine osmolality remaining well below isosmotic urine. Compared to that in patients with partial nephrogenic DI, patients with partial central DI usually achieve a urine osmolality of 300 mOsm/kg or higher after a vasopressin injection [, ]. After the vasopressin challenge test, he still complained with remained thirst (visual analogue scale 4). We treated the patient with hydrochlorothiazide (25 mg/day), and his urine output gradually decreased to within the normal range. After treatment of hydrochlorothiazide, his serum sodium level recovered to between 138 and 142 mEq/L. The patient recovered uneventfully and underwent treatment for Castleman’s disease in our hematology department.