SUMMARY: It has been suggested that haemodialysis adequacy is greater dialysing against a 3 mmol/L potassium dialysate concentration than against a 1 mmol/L potassium concentration. As most dialysis patients dialyse against 1 or 2 mmol/L potassium, the dialysis adequacy at these two potassium concentrations was compared. Ten stable haemodialysis patients were randomly assigned to dialyse against 1 mmol/L potassium dialysate (K1) followed by 2 mmol/L potassium dialysate (K2) or vice versa. All other dialysis parameters were held stable. The mean urea reduction ratio was 68.3 ± 6.2 using K1 and 69.5 ± 6.4 using K2 (P < 0.05 using Wilcoxon for paired data). The Kt/V, however, did not differ (1.39 ± 0.23 for K1 and 1.41 ± 0.23 for K2). The urea rebound was also not different between K1 and K2, with a trend to higher rebound using K2. The percentage rebound in urea was 6.0 ± 2.5 for K1 and 7.1 ± 2.8% for K2. In this setting, K2 dialysate offered no advantage in terms of urea rebound or Kt/V. Based on previously published data, a dialysate potassium concentration of 3 mmol/L may be required to achieve significant benefit in terms of dialysis adequacy.