Dialysis prescription commonly exceeds the delivered dialysis dose. Tunneled hemodialysis catheters (PC) may provide less dialysis than arteriovenous fistula (AVF) and polytetrafluoroethylene grafts (GG), but the impact of access type on the discrepancy (deltaHD) between dialysis prescription and dose is unknown. This study investigates the relationship between deltaHD and vascular access type. Fifty three chronic hemodialysis patients in our unit were prospectively studied for 3 weeks with measurement of delivered single pool and prescribed Kt/V(urea). There were 25 patients with AVF, 17 with GG, and 11 with PC. Demographic characteristics did not significantly differ between groups. Mean prescribed Kt/V(urea) was 1.73 +/- 0.26, and mean delivered Kt/V(urea) was 1.61 +/- 0.26. For 10 of 53 (19%) patients, dialysis delivery was at least equal to that prescribed, and this proportion did not differ between access types. Forty six of fifty three patients (86.7% of all patients) received Kt/V(urea) > 1.3, with no difference in this proportion between access types: AVF 22 of 25 (88.0%), GG 16 of 17 (94.1%), PC 8 of 11 (72.7%). Surprisingly, prescription times for patients with PC (3.6 +/- 0.3 hr) were significantly shorter than for those with AVF (3.9 +/- 0.3 hr) and GG (3.9 +/- 0.3 hr) (p = 0.02), perhaps indicating physician bias toward patients with tunneled catheters. In summary, access type was not a significant predictor of deltaHD, although patients with arteriovenous access tended to receive more dialysis than those with tunneled catheters. While a large proportion of patients received less dialysis than prescribed, the high levels of delivered Kt/V(urea) indicate that adequate dialysis is possible even in patients who must use tunneled catheters.