An experimental study of beer microfiltration has been carried out on ceramic membranes with the eventual aim of carrying this process through to the commercial scale. Enhancement of surface hydrodynamics through flow pulsation had little impact on flux suggesting that pore blocking by in-depth adsorption/deposition was the dominant factor and this was indeed found to be so. The nature of the foulants has been determined by studying the filtration rates of beer treated with various enzymes which degrade potential foulant species. Specific classes of carbohydrates and minerals have been identified as foulants. In particular, pentosans (carbohydrates composed of 5 numbered sugar rings) make a major contribution. A multi-stage backflush programme was developed and optimized in an attempt to achieve maximum pore clearance with minimal use of permeate and time. Moreover, when backflush (BF) was employed, staged increases in trans-membrane pressure had a more positive impact on flux improvement. The effect of membrane pore size on product quality and flux was also investigated in this work. Use of the BF programme achieved a flux improvement of 400%.