Although peritoneal dialysis (PD) is a widely accepted form of renal replacement therapy (RRT), concerns remain regarding the bioincompatible nature of standard PD fluid. In order to evaluate whether a newly formulated fluid of neutral pH, and containing low levels of glucose degradation products (GDP), resulted in improved in vivo biocompatibility, it was compared in a clinical study to a standard PD fluid. In a multicenter, open, randomized, prospective study with a crossover design and parallel arms, a conventional, acidic, lactate-buffered fluid (SPDF) was compared with a pH neutral, lactate-buffered, low GDP fluid (balance). Overnight effluent was collected and assayed for cancer antigen 125 (CA125), hyaluronic acid (HA), procollagen peptide (PICP), vascular endothelial growth factor (VEGF), and tumor necrosis factor alpha (TNFalpha). Serum samples were assayed for circulating advanced glycosylation end products (AGE), N(epsilon)-(carboxymethyl)lysine (CML), and imidazolone. Clinical end points were residual renal function (RRF), adequacy of dialysis, ultrafiltration, and peritoneal membrane function. Eighty-six patients were randomized to either group I starting with SPDF for 12 weeks (Phase I), then switching to "balance" for 12 weeks (Phase II), or group II, which was treated vice versa. Seventy-one patients completed the study with data suitable for entry into the per protocol analysis. Effluent and serum samples, together with peritoneal function tests and adequacy measurements, were undertaken at study centers on three occasions during the study: after the four-week run-in period, after Phase I, and again after Phase II. In patients treated with balance there were significantly higher effluent levels of CA125 and PICP in both arms of the study. Conversely, levels of HA were lower in patients exposed to balance, while there was no change in the levels of either VEGF or TNFalpha. Serum CML and imidazolone levels fell significantly in balance-treated patients. Renal urea and creatinine clearances were higher in both treatment arms after patients were exposed to balance. Urine volume was higher in patients exposed to balance. In contrast, peritoneal ultrafiltration was higher in patients on SPDF. When anuric patients were analyzed as a subgroup, there was no significant difference in peritoneal transport characteristics or in ultrafiltration on either fluid. There were no changes in peritonitis incidence on either solution. This study indicates that the use of balance, a neutral pH, low GDP fluid, is accompanied by a significant improvement in effluent markers of peritoneal membrane integrity and significantly decreased circulating AGE levels. Clinical parameters suggest an improvement in residual renal function on balance, with an accompanying decrease in peritoneal ultrafiltration. It would appear that balance solution results in an improvement in local peritoneal homeostasis, as well as having a positive impact on systemic parameters, including circulating AGE and residual renal function.