With the increased awareness of the problems associated with the growth dependent analysis of bacterial populations, direct optical detection methods such as flow cytometry have enjoyed increased popularity over the last few years. Among the analyses discussed here are: (1) Bacterial discrimination from other particles on the basis of nucleic acid staining, using sample disaggregation to provide fast reliable enumeration while minimizing data artefacts due to post sampling growth; (2) Determination of basic cell functions such as reproductive ability, metabolic activity and membrane integrity, to characterise the physiological state or degree of viability of bacteria; and (3) The use of single cell sorting onto agar plates, microscope slides or into multi-well plates to correlate viability as determined by cell growth with fluorescent labelling techniques. Simultaneous staining with different fluorochromes provides an extremely powerful way to demonstrate culture heterogeneity, and also to understand the functional differences revealed by each stain in practical applications. Analysis of bacterial fermentations showed a considerable drop (20%) in membrane potential and integrity during the latter stages of small scale (5L), well mixed fed-batch fermentations. These changes, not found in either batch or continuous culture fermentations, are probably due to the severe, steadily increasing stress associated with glucose limitation during the fed-batch process, suggesting 'on-line' flow cytometry could improve process control. Heat injured cells can already show up to 4 log of differences in recovery in different pre-enrichment media, thus contributing to the problem of viable but non-culturable cells (VBNC's). Cytometric cell sorting demonstrated decreasing recovery with increasing loss of membrane function. However, a new medium protecting the cells from intracellular and extracellular causes of oxidative stress improved recovery considerably. Actively respiring cells showed much higher recovery improvement than the other populations, demonstrating for the first time the contribution of oxidative respiration to intracellular causes of damage as a key part of the VBNC problem. Finally, absolute and relative frequencies of one species in a complex population were determined using immunofluorescent labelling in combination with the analysis of cell function. The detail and precision of multiparameter flow cytometric measurements of cell function at the single cell level now raise questions regarding the validity of classical, growth dependent viability assessment methods.