In situ hybridization with a fluorescently labeled 16S rRNA-targeted probe was examined using Rhodopseudomonas palustris as a model organism, which had been grown at different rates and under different conditions of growth and starvation. The specific growth rate did not affect the percentage of hybridized cells in aerobically grown R. palustris cultures. However, significant changes in the percentage of hybridized cells occurred during extended periods of starvation. These changes were observed both in batch cultures grown and starved aerobically in the dark, and in cultures grown phototrophically and starved anaerobically in the dark. Aerobic growth in batch culture and subsequent starvation resulted in a complete lack of detectable hybridization after 20 days of starvation. In contrast, even after 30 days of starvation, 50% of all cells were still detectable in cultures grown aerobically at growth rates <0.06 h(-1) and then starved aerobically in the dark. The same was true for phototrophically grown cells that were starved anaerobically in the light. During starvation there was a clear, though non-linear, positive correlation between the percentage of hybridized cells and the RNA content. In contrast, no direct correlation was observed between the number of hybridized cells in a culture and the viability of this culture. Thus, in habitats with growing, non-growing, and starving bacteria, data on quantitative detection of populations based on 16S rRNA-targeted probing should be used with extreme caution as the detectability of the individual cells is strongly influenced by their physiological history and current physiological state.