The growth efficiency of freshwater bacteria was examined in continuous cultures. One series of experiments was carried out using generation times from 50 to 200 hours and aged, normal, and enriched media, all of natural origin. Another series of experiments examined the bacterial growth efficiency during the growth season in eutrophic Frederiksborg Slotssø, in relation to changes in the planktonic communities and to factors controlling the bacterial incorporation of (3)H-thymidine. Attachment of bacteria to the inner surfaces of the experimental flasks was examined using various types of bottles, adding glass tubes to the bottles, and measuring (3)H-thymidine incorporation and direct cell counts of attached and free-living bacteria. Attachment of bacteria varied, and in one example up to 36% of the thymidine incorporation was by attached bacteria after 4 days. It was calculated that 36% of attached bacteria caused an underestimation of the growth efficiency of 11%. The mean growth efficiency tended to decrease with generation time using enriched medium (47 to 19%) and aged medium (35 to 12%), and tended to decrease with medium quality (enriched > normal > aged media) from 37% to 27%. The only significant difference in growth efficiency occurred in relation to generation time, in samples with enriched medium (unpaired t-test, P < 0.05). The overall mean value for all generation times and media was 30% (SEM = 3%, n = 24). From April to October, the growth efficiency was determined 5 times in samples from Frederiksborg Slotssø. The overall mean value was 31% (SEM = 3%, n = 30), and there was no significant change in the growth efficiency during the period measured. In June, three bioassay experiments revealed that carbon limitation controlled bacterial incorporation of (3)H-thymidine, whereas additions of phosphate and nitrate did not change the incorporation rates. The narrow range of growth efficiencies obtained in this study (mean 31%, SEM = 2%, n = 54) suggests that changes in substratequality in the media applied and in the eutrophic samples examined causes only subtle changes in the growth efficiency.