Solid phase extraction (SPE) procedures were used to determine the rec
overies of herbicides typically used in containerized ornamental plant
production from water samples. Recoveries from Cls cartridges and dis
ks were compared for each of 12 herbicides with variations in elution
solvent and volume of elution solvent tested. Recoveries for nine of t
he herbicides from the cartridges and disks using acetone as an elutio
n solvent were not affected by SPE matrix. Fluazifop recovery was grea
ter with the disks, while napropamide and oxadiazon recoveries were gr
eater with cartridges. Both cartridges and disks yielded low recoverie
s (23 to 47%) of benefin and prodiamine. Changing the elution solvent
from acetone to acetonitrile resulted in 10% improvement for the recov
ery of benefin and a three- to four-fold increase in recovery of prodi
amine. Acetonitrile decreased recoveries of napropamide, oryzalin, oxa
diazon, oxyfluorfen, and pendimethalin from cartridges. For the disks,
oxyfluorfen, prodiamine, and trifluralin had increased recovery, whil
e fluazifop, oxadiazon, and simazine had decreased recovery,vith aceto
nitrile as the elution solvent. Increasing the amount of acetone eluti
ng solvent increased the recovery of prodiamine and oxyfluorfen while
decreasing the recovery of fluazifop, pendimethalin, simazine, and tri
fluralin. Binding capacities of oryzalin on cartridges and disks avera
ged 13.2 and 7.8 mg, respectively. The advantage of the disk lies in t
he greater volume of water that can be processed, while the higher cos
t and greater variability are disadvantages. Cartridge extraction yiel
ded good recoveries with lower standard deviations, and used less orga
nic solvent. Selection of an SPE extraction method depends upon the he
rbicides under evaluation, expected levels, and the water volume being
processed. Both SPE techniques offer advantages over traditional liqu
id-liquid extraction methods such as reduced requirements for organic
solvent and sample preparation.