A recently developed magnetic-based ELISA test kit forthe herbicide metribuzin [4-amino-6-(1,1-dimethylethyl)-3-(methylthio)-1,2,4-triazin-5(4
H)-one] was evaluated foritsreproducibility, accuracy, and comparability to resultsfroma gas chromatography/mass spectrometry (GC/MS)technique for the determination of metribuzin in streamwater. Replicated metribuzin determination in streamwater showed good daily reproducibility (CV < 10%)exceptat low concentrations (<0.68
![](/images/entities/mgr.gif)
g L
-1). Althoughthe assaycorrelated well (
r = 0.954) with GC/MS results,thecorrelation was improved (
r = 0.990) whenconcentrations<0.68
![](/images/entities/mgr.gif)
g L
-1 were omitted. Predicted accuracy bythemetribuzin assay was tested in two matrices [stream anddeionized (DI) water]. Deionized water matrixexhibitedexcellent agreement between the theoretical and predictedmetribuzin levels, indicating a high degree of accuracy.Predicted metribuzin concentrations were slightlyelevatedcompared to theoretical concentrations in stream samples;however, no significant difference (
p < 0.05) wasfoundbetween the two sample matrices. This ELISA was foundto tolerate an excessive amount of several chemical substances routinely found in environmental samples as wellas a wide range of pH levels. Overall, the assayillustratedthe ability to efficiently and accurately predictconcentrations of metribuzin in stream water and demonstrated theutility of the ELISA technique as a screening tool for thedetermination of metribuzin in stream samples.