The benefits of daily haemodialysis (DHD) compared to conventional three times per week haemodialysis (CHD) have been described in a number of observational studies. Most of these previous studies however have not been performed with rigorous controls.We performed a crossover study following an A-B-A design: phase A was 4 weeks of thrice weekly dialysis, 3-4 h per treatment (CHD); phase B was 8 weeks of six times/week dialysis, each session being one-half of the usual time (DHD) and phase A with 4 weeks of thrice weekly dialysis (CHD) was repeated. Patients characteristics: n=12, six males; age 52+/-18 years, six diabetics.Weekly single-pool Kt/V, equilibrated Kt/V and standard Kt/V of urea, and beta-2-microglobulin clearance values were greater during DHD. Eight of 12 patients who completed the study reported symptomatic benefits from DHD that partially or completely disappeared during the second period of CHD. Quality of life of patients improved during DHD. Three patients had problems with arteriovenous access during DHD. Average blood pressure was lower during DHD (systolic 139.5+/-22.7 mmHg) compared to the initial (147.7+/-21.4 mmHg, P<0.001) and last (146.4+/-20.0 mmHg, P<0.005) CHD periods. No significant changes in predialysis haemoglobin and the serum concentration of albumin, phosphate, beta-2-microglobulin or B-type natriuretic peptides (BNP) were observed, although BNP trended downward during DHD and returned to baseline level during the second period of CHD. The dose of erythropoietin did not change significantly. Patient compliance with the dialysis schedule was lower during DHD. Dialysis staff perceived an increased workload but felt that the patients benefited medically from DHD.The results of this cross-over study suggest symptomatic benefits and decrease in blood pressure, but there are potential problems with compliance and vascular access during DHD.