A method for bioprocess monitoring using material balances with the aid of fuzzy identification of the culture state is proposed. Fuzzy inferencing was applied to estimate the varying material balances based on current culture data, which was then used to calculate the biological parameters backward from the CO2 evolution rate (CER) on-line. This was possible even though the metabolic balance altered as the fermentation progressed due to the changing culture conditions. Using this system, the cell, glucose and acetate concentrations, and the specific growth rate in a batch culture of recombinant E. coli, in which the material balances and cell yield changed in response to changes in the culture pH as cultivation proceeded, were successfully monitored based on the CER.