Coadhesion between oral microbial pairs is an established factor in the spatiotemporal development and prevalence of mixed-species communities in early dental plaque in vivo. This study compares removal and de novo adhesion of pairs of coadhering and non-coadhering oral actinomyces and streptococci by sonic brushing on salivary pellicles in a non-contact mode as a function of the distance between the brush and the pellicle surface in vitro. First, actinomycetes were adhered to a pellicle surface, after which streptococci suspended in saliva were allowed to adhere. Removal was examined by non-contact, sonic brushing with a wetted brush on a either a wetted or a substratum immersed to a depth of 7 mm. After brushing, de novo adhesion of streptococci to brushed pellicles was studied. For coadhering and non-coadhering pairs, 34% and 9%, respectively, of the adhering bacteria were involved in aggregates comprising more than 10 organisms. Non-contact, sonic brushing removed up to 99% of the adhering bacteria, regardless of the state of immersion of the substratum. Bacterial removal decreased with increasing distance of up to 6 mm between brush and pellicle surface. For the non-coadhering pair, subsequent exposure of pellicles to a streptococcal suspension yielded about 6% of bacteria involved in large aggregates. Alternatively, de novo adhesion of the coadhering streptococcal strain to pellicles brushed on the wetted substratum yielded 31% of bacteria involved in large aggregates, but after brushing the immersed substratum only 12% of the adhering bacteria were found in large aggregates. It is concluded that non-contact sonic brushing, under immersion, removes high percentage of adhering bacterial pairs up to a distance of 6 mm between the brush and the pellicle surface. However, non-contact, sonic brushing with only a thin wet film on the substratum may leave footprints to which streptococci preferentially adhere.