Sustained oscillations of biomass, substrate and product concentrations are routinely observed in continuous cultures withZymomonas at low growth rate and high ethanol concentrations. After an extensive work, Jöbses et al. (1985–1987)^5–7 proposed a two-compartment model to represent the phenomenon. However, the strong fluctuations of cell viability observed duringZymomonas cultures suggest that viability is a major parameter which unfortunately has never been completly taken into account. If a structuration of the cell population as viable, dead, and non viable cells (unable to divide but still able to produce ethanol) is introduced into a simple mathematical model, any situation from completly stable to completly unstable continuous cultures can be described. Experimental observations suggest that this approach may represent the biological reality more closely than the structured model proposed by Jöbseset al.