Patients with diabetes and renal insufficiency (Db/ESRD), a group subject to accelerated atherosclerosis exhibit marked increases in the levels of circulating, glycation-derived reactive substances, termed advanced glycation endproducts (AGEs). These products have bean previously shown to react covalently with apoliprotein B (ApoB) to form AGE-ApoB, a modification that results in delayed low density lipoprotein (LDL) clearance and possibly to dyslipidemia. Because the effect of hemodialysis on AGE removal was shown to be unsatisfactory, based on single intradialytic studies, we examined the effect of long-term hemodialysis therapy on serum AGE-ApoB levels, as well as on total serum ApoB of 25 Db/ESRD patients treated by two types of hemodialysis filters, the Fresenius Inc. F8, as the low flux (LF), or high-flux polysulfone AN69 (HF) for two months using an AGE-specific ELISA. At the end of eight weeks, circulating AGE-ApoB levels were reduced significantly (by 35%) from baseline (P = 0.039) in patients treated by HF compared to a modest 16% reduction noted in patients treated by LF (P = 0.05) N = 12, (P = 0.047). Of note, total plasma ApoB was reduced by 27% from baseline (P = 0.02) in patients treated by HF compared to a 6% reduction noted in those treated with LF (P = 0.8). In vitro comparison of AGE mass balance, and mass adsorption by the different filters revealed that the higher efficiency of HF filter was due to greater adsorption. The association of reduced AGE-ApoB levels with a decrease in total circulating ApoB by HF and not by LF dialysis suggests: (1) a causal link between AGE clearance and dyslipidemia in diabetic ESRD, and, (2) that more efficient modes of renal replacement treatment and AGE removal could significantly benefit clinical outcome.