In this work the feasibility of the recovery of a commercial sample of citrus pectin from model solutions using a temperature- and pressure-controlled bench-top ultrafiltration (UF) plant equipped with a ceramic tubular, 20 kDa nominal molecular weight cut-off (NMWCO), UF membrane module, was assessed. Several total recycle tests allowed the assessment of the effects of transmembrane pressure difference (Delta P), feed superficial velocity (v(s)) and solute concentration (C-BR) in the ranges of 0.4-4.2 bar, 4-6 m s(-1), and 2.2-30.4 kg m(-3), respectively, on the permeation flux under a constant process temperature of 50 degrees C. As c(BR) increased from about 2.2 to 10 kg m(-3) the limiting permeation flux (J(P infinity)) almost linearly decreased from 125-195 to about 25-40dm(3) m(-2) h(-1), these values depending on vs. For CBR rising from similar to 10 to 31 kg m(-3), J(P infinity) was almost constant or slightly tended to decline. The change in slope of the plot J(P infinity)-vs-log(C-BR) was attributed to the transition from turbulent to laminar regime. Two dimensionless empirical regressions among the modified Sherwood, Reynolds and Schmidt numbers, valid in the laminar or turbulent flow regime, enabled the estimation not only of the J(P infinity) values detected here in the total recycle mode, but also of the permeation fluxes observed by Pritchard et al. [M. Pritchard, R. Field, J.A. Howell, The ultrafiltration of viscous liquids, J. Membr. Sci. 102 (1995) 223-235] using a polyethersul phone, 65-kDa NMWCO, tubular module operating at 45 degrees C, AP = 2.5 bar and different feed superficial velocities (1.3-2.7 m s(-1)) with about 21% deviation. Moreover, such relationships yielded satisfactory simulation of two other independent validation tests performed in the batch mode with an average percentage error between the theoretical and experimental fluxes of 15%. (C) 2008 Elsevier B.V. All rights reserved.