The objective of this research was to evaluate the potential of using Fourier-Transform Infrared Microspectroscopy (ATR-IRMS) combined with multivariate analysis as a rapid method for studying the efficiency of cleaning protocols applied to reduce or eliminate membrane fouling after membrane emulsification. A coarse emulsion was prepared using 10% (v/v) sunflower oil and 90% (v/v) water phase containing 1% (w/v) whey protein. The coarse emulsion was passed through a nylon membrane (0.8 mu m pore size) by applying 900 kPa N(2) pressure. Five cleaning protocols combining different concentrations of Tween 20 (2, 3 and 4%) and backflush N(2) pressure (150, 500 or 700 kPa) were used to clean fouled membranes. New, fouled and clean membranes were dried out and analyzed by attenuated total reflectance using a FT-IR microspectroscopy in the mid-infrared region (4000-800 cm(-1)). Spectral data were used to create soft independent modeling of class analogy models (SIMCA) to differentiate among new, fouled and clean membranes. SIMCA models exhibited clusters that permitted clear differentiation among the cleaning protocols tested, identifying as the most effective treatments when 3 and 4% of Tween 20 and 700 and 500 kPa of N(2) pressure was applied, respectively. (C) 2011 Elsevier B.V. All rights reserved.