This study compared removal of pairs of co-adhering and non-co-adhering oral actinomyces and streptococci from salivary pellicles by manual, rotating/oscillating electric and sonic toothbrushes, applying weights up to 240 g. First, actinomyces were allowed to adhere to a pellicle in a parallel plate flow chamber, after which streptococci suspended in saliva were perfused through the chamber at 33°C. On average, 34–39% of the adhering bacteria were adhering as single organisms. For co-adhering and non-co-adhering pairs, 33 and 10% of the adhering bacteria were involved, respectively, in aggregates comprising more than 10 organisms. Brushing by hand removed 82% at low weight (40 g), which was less than by electric (93%) or sonic (92%) brushing, while for all modes of brushing bacterial removal increased with increasing weight to 95–99%. For a non-co-adhering pair, subsequent exposure of brushed pellicles to a streptococcal suspension yielded only 2–16% of bacteria involved in large aggregates, regardless of the mode of brushing. For the co-adhering pair, however, de novo streptococcal adhesion to hand-brushed pellicles yielded 34–57% of bacteria involved in large aggregates, while electric and sonic brushing left 22–35% of the bacteria involved in large aggregates. De novo streptococcal adhesion for the co-adhering pair increased with increasing weight for the electric and sonic brush in contrast to the manual brush. Since a strong influence of co-adhesion is evident in de novo streptococcal adhesion, despite nearly complete removal of all actinomyces, these observations suggest that the three modes of brushing leave footprints to which streptococci preferentially adhere.