A customized stirred-tank biofilm reactor was designed for plastic-composite supports (PCS). In repeated-batch studies, the PCS-biofilm reactors outperformed the suspended-cell reactors by demonstrating higher lactic acid productivities (2.45 g l(-1) h(-1) vs 1.75 g l(-1) h(-1)) and greater glucose consumption rates (3.27 g l(-1) h(-1) vs 2.09 g l(-1) h(-1)). In the repeated fed-batch studies, reactors were spiked periodically with concentrated glucose (75%) to maintain a concentration of approximately 80 g of glucose l(-1) in the bioreactor. In suspended-cell fermentations with 10 g of yeast extract (YE) l(-1) and zero, one, two, and three glucose spikes, the lactic acid productivities were 2.64, 1.58, 0.80, and 0.62 g l(-1) h(-1), respectively. In comparison, biofilm reactors with 7 g of YE l(-1) and zero, one, two, and three glucose spikes achieved lactic acid productivities of 4.20, 2.78, 0.66, and 0.94 g l(-1) h(-1), respectively. The use of nystatin (30 U ml(-1)) subdued the contaminating yeast population with no effect on the lactic acid productivity of the biofilm reactors, but it did affect productivity in the suspended-cell bioreactor. Overall, in repeated fed-batch fermentations, the biofilm reactors consistently outperformed the suspended-cell bioreactors, required less YE, and produced up to 146 g of lactic acid l(-1) with 7 g of YE l(-1), whereas the suspended-cell reactor produced 132 g l(-1) with 10 g of YE l(-1).