The impact of glufosinate-ammonium and bialaphos on the phytoplankton community in a shallow eutrophic lake ecosystem was investigated using in situ enclosures. Flow cytometry was used to size phytoplankton cells and sort them as live or dead, depending upon their chlorophyll a autofluorescence intensity. Flow cytometric analyses provided significantly (p Ͻ 0.001) higher abundance estimates as compared to conventional microscopic analyses. At the highest treatment levels (10 mg/L), both herbicides caused a significant but transient reduction in live phytoplankton cells (days 3-14), which was particularly apparent in the small (1-2 and 2-3 m) classes. Transient impacts on phytoplankton live cell abundances were mirrored by depression in dissolved oxygen content in the treated enclosures. At an application rate of 10 mg/L, abundance of smaller phytoplankton in the bialaphos-treated enclosures recovered more rapidly (14 d) than those in the glufosinate-ammonium-treated enclosures (49 d). For days of maximal impact, estimated median effect concentrations (EC50) for reduction of phytoplankton abundance ranged from 2.5 to 3.4 mg/L for glufosinate-ammonium and 3.3 to 8.1 mg/L for bialaphos, whereas estimates of concentration inducing 20% reductions in abundance (EC20; 0.9-1.2 and 1.6-4 mg/L, respectively) more closely approximated the expected environmental concentration (1 mg/L), assuming direct overspray into water bodies of 15-cm depth.