The proportion of vital as compared with dead Streptococcus sanguinis cells attached to glass surfaces was monitored and related to varying proportions of planktonic vital as compared with dead Strep. sanguinis cells. In a flow chamber with six parallel-mounted glass plates, Strep. sanguinis was suspended in pretreated sterile human saliva. Deposition of Strep. sanguinis took place, with a proportion of vital sanguinis streptococci in saliva (%VSs) of 90%, 45% or 22.5%. After exposure times of 30, 60, 90, 120 and 240 min, adherent microorganisms were labelled with two fluorescence stains to differentiate between vital and dead bacteria. Proportions of vital attached streptococci (%VSa) were determined microscopically. Dead bacteria were detected on all glass plates. The %VSa at 30 min and 60 min was significantly lower than the baseline %VSs. During the course of a single run the %VSa frequently increased after either 30, 60 or 90 min without exceeding the %VSs at 4 h. %VSs was the only variable exerting a significant effect on %VSa at 30 and 60 min. It is suggested that during the initial events of microbial attachment the dead rather than vital Strep. sanguinis cells attach preferably to solid surfaces.