Whether salt or water intake is the primary cause of interdialytic weight gain (ΔW) has important implication for the design of measures to prevent large ΔW. In 17 hemodialysis patients dialyzed against a bath containing 140 mmol/L of sodium, monthly predialysis serum sodium was compared with post dialysis serum sodium. A decrease in serum sodium in the interdialytic period would indicate that primary water consumption accounts for at least part of the ΔW. Interdialytic sodium intake, isotonic fluid gain (ΔWisotonic) and net pure water gain (ΔWH2O) were calculated by balance formulae. Serum sodium concentration was corrected in diabetic subjects to the value corresponding to euglycemia (100 mg/dl). Estimated interdialytic sodium intake was compared with the prescribed sodium intake and, in seven subjects, to sodium intake estimated from dietary records. Results for nondiabetic subjects (N = 9): [Na]post 139.3 ± 1.9 mmol/L, [Na]pre 140.1 ± 2.1 mmol/L (NS), ΔW 1.15 ± 0.55 L/24 hr, Δ Wisotonic 1.33 ± 0.57 L/24 hr, ΔWH2O −0.20 ± 0.58 L/24 hr, estimated sodium intake 206 ± 75 mmol/24 hr, prescribed sodium intake 121 ± 29 mmol/24 hr (p = 0.028). Results for diabetic subjects (N = 7): [Na]post140.1 ± 2.5 mmol/L, [Na]pre 137.7 ± 3.1 mmol/L (p < 0.01), ΔW 1.26 ± 0.38 L/24 hr, ΔWisotonic 0.59 ± 0.63 L/24 hr, ΔWH2O 0.66 ± 0.39 L/24 hr, estimated sodium intake 160 ± 81 mmol/24 hr, prescribed sodium intake 124 ± 30 mmol/24 hr (NS), glycosylated hemoglobin 9.7 ± 2.8% (normal, 4.1–5.7%). In seven subjects, estimates of sodium intake from balance formulae (233 ± 113 mmol/24 hr) were not different from estimates from dietary records (212 ± 87 mmol/24 hr). Sodium intake accounted for all the interdialytic weight gain in nondiabetic subjects. In diabetic patients, only approximately half of the interdialytic weight gain was accounted for by sodium intake. The other half was due to pure water gain, probably caused by hyperglycemia.