Hemodialysis vascular access-related problems account for most hospitalizations in chronic hemodialysis patients. Although some co-morbid risk factors for early fistula failures have been described, a great deal of unknown exists as to why access survival is favorable in some patients. In this longitudinal study, fistulae patency and thrombosis episodes were monitored from placement date in three groups of end-stage renal disease (ESRD) patients who have been on dialysis for > or =90 days. Thirty-six patients (29 male; 80%) with a mean age of 42+/-2 years were monitored. The groups consisted of eight patients with biopsy-confirmed focal segmental glomeruloscierosis (FSGS), 13 with acquired immunodeficiency syndrome-related nephropathy (human immunodeficiency virus [HIV]), and 15 with hypertensive ESRD (hypertensive nephrosclerosis [HTN]) who served as controls. Diabetics and patients aged > or =64 years were excluded. Twenty-five of 36 (69%) fistulae were prosthetic (AVG), while 11 (31%) were native (AVF). The FSGS group was more likely to have an AVG (87.5%), while 54% of the HIV group had an AVG. The thrombosis event rate was significantly greater among the FSGS patients (3/patient-year) than the HIV (0.15/patient-year) and HTN (0.5/patient-year) patients (P < 0.0001 and P < 0.002, respectively). The mean thrombosis-free duration for both AVG and AVF among the HIV and HTN groups were 318.5+/-17 days and 311.7+/-22.5 days, respectively. These were significantly greater than in the FSGS group (26.5+/-7 days; P < 0.0001). The cumulative 1-year patency rate for AVG among the HIV and HTN groups was 85% and 65%, respectively, while that of the FSGS group was 0%. Kaplan-Meier hazard analysis showed that all groups were at risk of access thrombosis as time progressed, but the FSGS group had the highest risk of access thrombosis, which began from the date of placement and increased exponentially with time. The increased thrombosis rate among the patients in the FSGS group correlated with their weight (R = 0.8, P = 0.003) and pre-ESRD 24-hour urinary protein excretion (R = 0.9, P = 0.001). The HIV status appeared to confer enhanced hemodialysis access survival. This may be related to the high rate of native fistulae placement and favorable vascular reactivity to shear stress. Accelerated atherosclerosis and small caliber vessels may be responsible for the poor fistulae outcome among the FSGS group. More studies will be necessary to further explore these findings.