The relationship between flow cytometry data and epifluorescence microscopy measurements was assessed in bacterioplankton samples from 80 lakes to estimate bacterial biovolume and cell size distribution. The total counts of 4′,6′-diamidino-2-phenylindole-stained cells estimated by both methods were significantly related, and the slope of their linear regression was not significantly different from 1, indicating that both methods produce very similar estimates of bacterial abundance. The relationships between side scatter (SSC) and 4′,6′-diamidino-2-phenylindole fluorescence and cell volume (microscopy values) were improved by binning of the data in three frequency classes for each, but further increases in the number of classes did not improve these relationships. Side scatter was the best cell volume predictor, and significant relationships were observed between the SSC classes and the smallest (R2 = 0.545, P < 0.001, n = 80) and the largest (R2 = 0.544, P < 0.001, n = 80) microscopy bacterial-size classes. Based on these relationships, a reliable bacterial biomass estimation was obtained from the SSC frequency classes. Our study indicates that flow cytometry can be used to properly estimate bacterioplankton biovolume, with an accuracy similar to those of more time-consuming microscopy methods.Sampling and microscopy analyses were done within the framework of the European project EMERGE (EVK1-CT-1999-00032). Work at Innsbruck was supported by the Austrian Science Foundation (grant P-19245-BO3) and by a postdoctoral grant to M.F. from the European Commission (no. ENV4-CT98-5099).Peer reviewed