Redox potential was used to develop a stationary-phase fermentation of Candida tropicalis that resulted in non-growth conditions with a limited decline in cell viability, a xylitol yield of 0.87 g g(-1) (95% of the theoretical value), and a high maximum specific production rate (0.67 g g(-1) h(-1)). A redox potential of 100 mV was found to be optimum for xylitol production over the range 0-150 mV. A shift from ethanol to xylitol production occurred when the redox potential was reduced from 50 mV to 100 mV as cumulative ethanol (Y-ethanol) decreased from 0.34 g g(-1) to 0.025 g g(-1) and Y-xylitol increased from 0.15 g g(-1) to 0.87 g g(-1) (alpha=0.05). Reducing the redox potential to 150 mV did not improve the fermentation. Instead, the xylitol yield and productivity decreased to 0.63 g g(-1) and 0.58 g g(-1) h(-1) respectively and cell viability declined. The viable, stationary-phase fermentation could be used to develop a continuous fermentation process, significantly increasing volumetric productivity and reducing downstream separation costs, potentially by the use of a membrane cell-recycle reactor.