B-phycoerythrin is a major light-harvesting phycobiliprotein in some marine algae. It is widely used as a fluorescent probe and analytical reagent, and could also be used as a natural dye in foods and cosmetics. In the present work, an expanded-bed methodology for large-scale recovery of B-phycoerythrin from the microalga Porphyridium cruentum is developed using Streamline-DEAE as adsorbent. The operation of expanded bed technology was first optimized on a small scale using a column of 15 mm internal diameter. The chromatographic parameters studied were the sample load, viscosity and expansion degree. The optimal conditions proved to be a sample load of 0.88 mg B-PE/mL Streamline DEAE, an expanded bed volume twice the settled bed volume and a sample viscosity of 1.068 mP. The expanded bed adsorption process was then scaled up 16 times by increasing the column diameter, while maintaining the sedimented bed height, linear flow rate and protein load at constant values. The success of the scale-up process was verified by determining the protein breakthrough capacity and product recovery. A comparative study was made of the performance of four columns of 15, 25, 40 and 60 mm internal diameters. The standard deviation of the breakthrough curves ranged from 0.81 to 0.76; the higher the internal diameter the lower the standard deviation. The yield of the EBA chromatography was in the range of 71-78%. These results show that small diameter columns can be effectively used for mimicking the behaviour in scaled up systems providing a useful tool for method scouting studies. The results also show that the large-scale assay (60 mm diameter column) was as efficient as the lab scale one (15 mm diameter column) in recovering B-phycoerythrin from the unclarified crude extract. Results show that expanded bed chromatography is a scaleable technology that allows large quantities of B-PE to be obtained without product loss. It maintains a high protein recovery while reducing both processing costs and times, when compared with previous methodologies. (c) 2006 Elsevier Inc. All rights reserved.