Given the considerable time and expense invested in a single bioprocess (fermentation) batch, variability and losses must be identified quickly. We propose that "sniffing" the odor of cultivation media and broth using instruments could provide a rapid and early indication of bioprocess performance. The human sensation of odor is related to the molecular composition of the vapor phase. The traditional approach to characterize volatile compounds has been sample extraction followed by GC-MS analysis. This approach is very tedious and requires some knowledge of the molecules involved. A new, alternate approach based on an "electronic nose" is now available which, like the human nose, can directly characterize the odor without reference to chemical composition. Here, an array of "conductive polymer" sensors with different chemical sensitivities produces a set of different responses to the same odor. The responses are analyzed mathematically, using pattern recognition techniques, to differentiate between different odors with a high level of sensitivity. In this report, we demonstrate the feasibility of using a commercially available electronic nose for the following applications : monitoring lot-to-lot variation in bioprocess medium ingredients, detecting microbial contamination early, and evaluating bioprocess performance during cultivation of microorganisms at inoculum and production stages.