Glucose or its polymer is usually added to dialysis solution for the development of sufficient ultrafiltration during peritoneal dialysis. The aim of the present study was to determine the influence of glucose and icodextrin on the transport of gentamicin and insulin from the mesothelial to the interstitial side of the peritoneal membrane. Transfer values are expressed as a coefficient of diffusive permeability, P, in centimeters per second. Each of the molecules was tested in 3 series of experiments using rabbit parietal peritoneum, a modified Ussing chamber, and a mathematical model of mass transport. First, transperitoneal transfers of gentamicin (0.040 g/dL) and insulin (0.1 g/dL) were analyzed in control conditions for 120 minutes. Then, transport parameters for gentamicin and insulin were separately determined before (15-60 minutes) and after (75-120 minutes or 75-130 minutes) the application of glucose (1.8 g/dL) or icodextrin (2 g/ dL) on the mesothelial side of the peritoneal membrane. Insulin transport was observed to be stable in the control series. Gentamicin transfer was not stable; its passage declined by 52% (p < 0.01) in the control series. The mean transfer parameters were 7.41 +/- 1.40 cm/s (x0.0001) over 15-30 minutes and 3.21 +/- 0.54 cm/s (x0.0001) over 75-130 minutes. Gentamicin transfer declined less in the series with glucose or icodextrin, by 21% (p < 0.04) and 30% (p < 0.05) respectively, than in the control series. For insulin, the mean P (+ standard error of the mean) was 0.15 +/- 0.02 cm/s (x0.0001) at the first hour of transfer and 0.14 - 0.02 cm/s (x0.0001) at the second. Glucose induced a nonsignificant intensification of insulin transport. Icodextrin increased insulin passage by 107% (p < 0.03). Osmotic and oncotic factors (glucose and icodextrin) both stabilize the transfer of gentamicin across the peritoneal membrane in vitro. Glucose polymer intensifies insulin transport from the mesothelial to the interstitial side of the peritoneum. Similar modifications might be observed in vivo during peritoneal dialysis or continuous intraperitoneal administration of insulin, influencing the efficiency of those treatments.