A bench-scale study was undertaken to evaluate aluminum electrocoagulation pretreatment for dead-end, constant pressure microfiltration of surface water. Electrochemical aluminum production quantitatively obeyed Faraday's law with a 3-electron transfer at nearly 100% efficiency resulting in Al(OH)(3) precipitation. X-ray diffractometry proved that Al(OH)(3) precipitates were predominantly amorphous. Enmeshment of colloids in amorphous precipitates (sweep flocculation) was the predominant colloid destabilization mechanism. Additionally, the floc zeta potential became less negative (-> 0) as more and more aluminum was added indicating the secondary role of adsorption of hydrolyzed aluminum species and charge neutralization in destabilization. Fouling of a commercially available polymeric microfiltration membrane following electrocoagulation pretreatment was found to (i) be lower at pH 6.4 compared with 7.5, (ii) decrease only up to an intermediate aluminum dosage, and (iii) exacerbate with increasing transmembrane pressure. In all experiments, cake filtration was the predominant flux decline mechanism for the entire duration of microfiltration. Increased fouling at higher transmembrane pressures (manifested as more rapid decline of the instantaneous flux normalized by the initial flux) was attributed to cake compressibility. Optical microscopy revealed that floc sizes increased monotonically with aluminum dosage even though fouling worsened at high electrocoagulant concentrations. This suggests that fouling was controlled by antagonistic effects of adding more and more aluminum coagulant; increasing total mass loading of colloidal foulants (higher total filtration resistance) and creating larger flocs (decreasing specific cake resistance). Significant reductions in microfilter fouling with aluminum electrocoagulation pretreatment suggests the next logical step of larger-scale testing under "real-world" conditions before recommending it for surface water treatment. (C) 2011 Elsevier B.V. All rights reserved.