Nuisance populations of egeria (Egeria densa Planchon) hampered the use of the 1093 ha Chickahominy Reservoir, Virginia, as a source of recreation and potable water. A mixture of diquat dibromide (6,7-dihydrodipyrido (l,2-a:2,l'c)pyrazinediium dibromide) and potassium endothall (7- oxa.bicylq;>, z,1 )heptane-2 ,J-dicarboxylic acid) was applied at 2.83 liters of each chemical per 0.4 surface ha. There were no measurable changes in water quality after treatment and macrophyte die-off with the exception of a general decrease in dissolved oxygen; no fish kills were observed. Both herbicides declined to very low levels in the water 3 days after treatment and were undetectable within 16 days after treatment. Concentrations of diquat in.egeria were twice as high and persisted two times longer than concentrations of endothall. No endothall was found in bottom sediments. Diquat persisted in hydrosoils for one year. No herbicides were found in fish muscle tissue. Diquat and the mixture were equally effective at all concentrations tested in laboratory experiments at controlling five species of aquatic plants. Toxicity of endothall was much less than the toxicity of diquat and the mixture. Quantitative plant sampling augmented by aerial infrared photography revealed that egeria had repopulated the shallow areas of the reservoir one year after treatment but only repopulated the deeper areas by 42% by two years after treatment. LYngbya sp. increased in quantity especially in the deeper areas following treatment. Floating artificial substrates were employed to monitor in situ effects of treatment on macroinvertebrates. No toxic effects due to treatment were noted. Weed removal also did not affect growth of bluegill sunfish (Lepomis macrochirus). Tissues of bluegills collected after treatment did not histologically appear to be affected by the herbicides. The Lc50 values for diquat and endothall separately and combined were determined for four species of fish. The toxicity of the mixture was set by the toxicity of diquat. Concentrations approximating those applied in the field (0.11 mg/liter diquat with 0.17 mg/liter endothall) did not alter bluegill respiration rate. Concentrations of the mixture as high as 11.0 mg/liter diquat and 17.0 mg/liter endothall tested over 96 h in bioassays did not cause histopathological changes in bluegill gill, kidney, or liver tissue or cause changes in hematocrit, total plasma proteins, plasma protein electrophorograms or white blood cell differentials. Goldfish exposed to diquat and endothall separately (maximum 11.0 mg/liter diquat and 17.0 mg/liter endothall) and combined (maximum 11.0 mg/liter diquat with 17.0 mg/liter endothall) had no alterations in plasma glucose, glucocorticoids, androgens, estrogens, hernatocrit, or white blood cell differential. No histopathological changes due to treatment were found in goldfish interrenal, thyroid, hernatopoietic or sensory tissues; gill lesions occured at maximum concentrations. The mixture field concentration did not cause avoidance in goldfish. Higher concentrations of diquat (1.1 mg/liter) and the mixture (1.1 mg/liter diquat with 1.7 mg/liter endothall) were avoided. Endothall was not avoided until concentrations were increased to 17.0 mg/liter. Anglers agreed with the use of herbicides to control aquatic weeds, and were willing to refrain from fishing while the treatment was carried out. Anglers believed that the Chickahominy weed control project increased fishing and boating enjoyment.