The purpose of the study was two-fold: to determine the nature of stainable deposits on periodontally diseased root surfaces subsequent to in vivo scaling and root planing procedures, and to quantify the distribution of residual plaque on instrumented root surfaces. Thirty molar and 30 nonmolar teeth which were condemned for periodontal or prosthetic reasons and had proximal probing depths of 4 to 7 mm were treated. Half of these were instrumented with I.U. curettes and the other half with an ultrasonic scaling device. Instrumentation was continued until the root surface felt hard and smooth to an explorer tip. The location of the gingival margin was recorded by notching the treated proximal surface with a No. 1/2 round bur. Twenty control teeth, 10 molar and 10 nonmolar, were extracted without instrumentation. Control and experimental teeth were irrigated with saline and stored in a 2.5% glutaraldehyde fixative solution until the time of assessment. All teeth were stained with a 0.5% solution of toluidine blue, and the amount of residual stained material and calculus was assessed under the stereomicroscope using an eyepiece fitted with a 10 X 10 optical grid. Stained deposits were marked by placing small V-shaped notches in the adjacent root surface as an aid to identification after the specimens were processed for scanning electron microscopic (SEM) examination. The nature of stained deposits on selected teeth was then characterized using the SEM. Treated root surfaces were also surveyed in detail to assess the quantity and extent of residual plaque deposits. The findings showed that although a large percentage of the treated proximal root surface may possess stainable deposits, these surfaces were often unexpectedly free of microbial organisms. In this study, the majority of stained deposits were composed of adherent fibrin and instrumentation debris. When bacterial plaque was present, it was usually found in small "mini-colonies" smaller than 0.5 mm across. Such findings cast doubt on the validity of using histologic and disclosing stains as an indicator for the presence of bacterial plaque immediately after instrumentation. Although only partially effective in removing subgingival calculus, both methods of instrumentation in this study appeared to be remarkably effective in bacterial debridement of subgingival root surfaces.