A circulating sodium pump inhibitor, released in response to volume expansion, may, by increasing intracellular sodium concentrations, be responsible for some of the features of the uraemic state. Haemodialysis, by correcting volume overload, would be expected to be associated with a decrease in intracellular sodium towards normal. The effects of a haemodialysis session on leukocyte (WBC) sodium content and transport have not been described and there are conflicting reports of the effects of haemodialysis on erythrocyte (RBC) sodium content and transport. We have measured sodium (NaWBC/RBC) and potassium content (KWBC/RBC), net ouabain-sensitive sodium flux rate (FR) and sodium flux rate constant (RC) before and after a standard haemodialysis session in 20 stable hospital haemodialysis patients. In leukocytes (n = 18), sodium (P = 0.078), FR (P = 0.006), and RC (P = 0.071) decreased over dialysis, whereas in erythrocytes sodium increased (P less than 0.001, n = 19) and RC declined (P = 0.002, n = 18). Although in opposite directions in RBC and WBC, the changes in sodium were toward normal in both cell types. Changes in intracellular sodium content and transport did not correlate with changes in measures of ECF volume or biochemical efficiency of dialysis. We conclude that haemodialysis does affect cell sodium content and transport, but in different ways in different cell types. There was no evidence that haemodialysis removed a sodium pump inhibitor. The use of RBC or WBC as 'model' cells to study sodium transport in uraemia is of questionable validity.