Purpose: Our aim was to examine dentists' confidence in their decisions on whether or not to administer antibiotic prophylaxis and to analyze the effects of factors that might explain the interindividual variation in confidence. Methods: Postal questionnaires in combination with telephone interviews were used. The response rate was 51% (101/200). Eight simulated patient cases with a medical history that might imply an increased risk of infection following three different dental procedures were described in the questionnaire. The dentists were asked to assess on a 100-millimetre visual analogue scale how confident they were in their decisions of whether or not to administer antibiotic prophylaxis. We also analyzed the extent to which the interindividual variation in dentists' confidence assessments could be explained by the medical conditions and/or the dental procedures (R2), and divided the dentists into different categories. A discriminant analysis was then used to examine differences in background variables (sex, age, years of professional experience, place of work) between the categories of dentists. Results: Overall, the dentists exhibited high confidence in their decisions, regardless of whether they administered antibiotic prophylaxis (P>0.05). There was a wide variation in R2-values (range 0.293-0.996), explained by the medical conditions and/or the dental procedures. Individual variations in confidence were explained by the medical conditions in 46% of the dentists (P<0.05), by the dental procedures in 8% (P<0.05), and by neither the medical conditions nor the dental procedures in 46% (P>0.05). Differences in background variables between the three categories of dentists were nonsignificant (P>0.05). Conclusions: Regardless of whether the dentists administered antibiotics, or whether their decisions were evidence-based, they exhibited high confidence. This may indicate that dentists are less likely to modify their behavior and that implementation of evidence into practice would be difficult. For nearly half the dentists, the individual variation in confidence could not be explained by the medical conditions or the dental procedures. This suggests that decisions might be intuitive and not based on adequate information. The analysis of background variables yielded no differences among the three categories, which suggests that other characteristics, perhaps personality, may explain the interindividual variation in confidence among dentists.