Inflammation is commonly associated with malnutrition and cardiovascular disease in end-stage renal failure patients. Anti-inflammatory properties of the isoflavones, a micronutrient component of soy, have been reported in several experimental models and disease conditions, but never in renal failure. We hypothesized that dietary soy isoflavones correct laboratory evidence of systemic inflammation in haemodialysis (HD) patients with underlying high blood levels of C-reactive protein (CRP). End-stage renal disease patients on chronic HD, with elevated CRP (>10.0 mg/l) were enrolled in this pilot study. The subjects were double-blind randomly distributed with 2 : 1 ratio to receive isoflavone-containing soy-based nutritional supplements (soy group) or isoflavone-free milk protein (control group) for 8 weeks. Serum isoflavone, inflammatory markers and nutrition markers were assessed at baseline and at the end of the treatment. Thirty-two subjects were enrolled. Fifteen subjects in the soy group and 10 in the control group completed the study; five dropouts were due to acute illness and two due to food intolerance. After intervention, blood isoflavone levels were 5- to 10-fold higher in the soy group than in the control group [e.g. median genistein (25-75th percentile): 337.9 (175.5-1007) nM in the soy group vs 41.4 (22.9-100.4) nM in the control group; P < 0.001]. However, the isoflavone levels ranged widely in the soy group (e.g. genistein: 33-1868 nM) and, depending on the individual compound, four to seven subjects had end-of-treatment levels that were not different from baseline. Variation from baseline of the individual serum isoflavone levels (Delta-isoflavone) and CRP displayed a strong inverse correlation in the soy group (R = -0.599, P < 0.02). In addition, Delta-isoflavone correlated positively with the variation of albumin (R = 0.522, P = 0.05) and insulin-like growth factor-1 (R = 0.518, P < 0.05). Group levels of CRP were not statistically different after intervention although a trend towards lower levels was noted in the soy group [18.2 (12.7-29.1) mg/l at baseline vs 9.7 (5.2-20.7) mg/l at week 8; NS] but not in the control group [20.6 (9.2-38.5) vs 17.6 (9.1-40.7) mg/l]. These data suggest the possibility of beneficial effects of isoflavone-rich soy foods on the inflammatory and nutritional status of HD patients with underlying systemic inflammation.