A very sensitive and specific analytical procedure for determining 20 acidic pesticides in aqueous environmental samples using pneumatically assisted electrospray (ES) LC/MS is presented. This procedure involves passing 1- and 4-L river water and drinking water samples, respectively, through a 1-g graphitized carbon black (GCB) extraction cartridge. By exploiting the presence of positively charged active centers on the GCB surface, isolation of the acidic pesticides from base/neutral species was made possible by differential elution. Recoveries of the analytes were higher than 85%, irrespective of the aqueous matrix in which they were dissolved. Adoption of the ion pair technique by addition of 0.1 mmoVL K 2 HPO 4 and 0.2 mmoVL tetrabutylammonium fluoride to the mobile phase allowed analytes to be analyzed as preformed ions. A conventional 4.6-mm-i.d. reversed-phase LC C-18 column operating with a 1 mL/min flow of the mobile phase was used for chromatographing the analytes. A flow of 30 μL/min of the column effluent was diverted to the ES source, while the rest of the mobile phase was delivered to a UV detector set at 220-nm wavelength. The effects of the nature of the ion pair forming agent and its counterion as well as its concentration in the mobile phase on the response of the ES/MS detector were investigated. The effects of varying the skimmer cone voltage on the production of diagnostic fragments and the response of the MS detector were also evaluated. For the analytes considered, the response of the mass detector was linearly related to the amount of the analytes injected between 2.5 and 200 ng. A certain variation of the ion signal for the analytes considered occurred after several hours of continuous use of the LC/ES/MS instrumentation. For routine use, then, analyte quantitation could be better performed by the UV trace method while entrusting unambiguous identification of the analytes to the MS detector. The limits of sensitivity (signal-to-noise ratio = 3) of the method for the pesticides considered in drinking water and surface water samples were estimated to be about 2-6 and 8-25 ng/L, respectively.