We compared the detection of bacteria and archaea in the coastal North Sea and at Monterey Bay, Calif., after fluorescence in situ hybridization (FISH) either with rRNA-targeted oligonucleotide probes monolabeled with the cyanin dye Cy3 (oligoFISH) or with fluorescein-labeled polyribonucleotide probes (polyFISH). During an annual cycle in German Bight surface waters, the percentages of bacteria visualized by polyFISH (annual mean, 77% of total counts) were significantly higher than those detected by oligoFISH (53%). The fraction of total bacteria visualized by oligoFISH declined during winter, whereas cell numbers determined by polyFISH remained constant throughout the year. Depth profiles from Monterey Bay showed large differences in the fraction of bacterial cells visualized by polyFISH and oligoFISH in the deeper water layers irrespective of the season. Image-analyzed microscopy indicated that the superior detection of cells by polyFISH with fluorescein-labeled probes in bacterioplankton samples was less a consequence of higher absolute fluorescence intensities but was rather related to quasi-linear bleaching dynamics and to a higher signal-to-background ratio. The relative abundances of archaea in North Sea and Monterey Bay spring samples as determined by oligoFISH were on average higher than those determined by polyFISH. However, simultaneous hybridizations with oligonucleotide probes for bacteria and archaea suggested that the oligoFISH probe ARCH915 unspecifically stained a population of bacteria. Using either FISH technique, blooms of archaea were observed in North Sea surface waters during the spring and summer months. Marine group II archaea (Euryarchaeota) reached >30% of total picoplankton abundances, as determined by polyFISH. We suggest that studies of pelagic microbial community structure using oligoFISH with monolabeled probes should focus on environments that yield detections > or =70% of total cell counts, e.g., coastal surface waters during spring and summer.