METHODS The prevalence and intensity of bacteremia of dental origin were examined in 207 children divided into four groups: a baseline with no surgical intervention (group I), after a single tooth extraction (group II), multiple tooth extraction (group III), and mucoperiosteal flap elevation (group IV). The bacterial isolates were grown using a broth culture (Bactec) and lysis centrifugation (Paediatric Isolator) techniques. Dental plaque deposits, gingivitis, spontaneous gingival bleeding and the presence/absence of a dental abscess were recorded and their relationship to bacteremia assessed. RESULTS The broth culture was positive for group I 11% of the time, group II for 43%, group III for 54%, and group IV for 43%. The Paediatric Isolator system was found to be a poor method for detecting bacteremia, having only one quarter the sensitivity of the broth culture technique. When organisms were isolated, the intensity of bacteremia ranged from 1 to 3400 colony forming units per milliliter (cfu/mL). Bacterial isolates were susceptible to most of the antibiotics recommended for antibiotic prophylaxis, but erythromycin, gentamycin, penicillin G, and teicoplanin were only 80% (or less) effective in their efficacy while chlorhexidine, amoxicillin, clindamycin, and vancomycin were between 92 and 100% effective. CONCLUSIONS The antibiotics commonly used for an oral and/or parenteral prophylaxis are likely to be effective on at least 80% of occasions with most of them effective on 100% of occasions.