It is well known that most periodontal pathogens are also endodontic pathogens. Little is known, however, about microflora of combined periodontal-endodontic lesions. Such coexisting pulpal and periodontal inflammation affecting the same tooth is relatively rare and can complicate the diagnosis and treatment planning of the involved tooth. Objectives: To investigate quantitatively and qualitatively five selected bacteria known as periodontal pathogens, in samples from combined periodontal-endodontic lesions and in a similar number of autologous samples from simple periodontal lesions. Materials and methods: Paired subgingival plaque and root canal samples from combined lesions and subgingival samples from 14 separate periodontal sites (reference sites) in 19 patients (12 women and 7 men; mean age 44±19.3 years) were collected using sterile paper points. The 52 plaque samples, 33 subgingival and 19 from root canals, were cultivated and processed for bacterial DNA extraction. Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Tannerella forsythensis, Prevotella intermedia and Treponema denticola were identified using conventional polymerase chain reaction (PCR). Heteroduplexes formation and analysis were used for identification of different P. gingivalis strains. Bacterial growth was assessed as the number of anaerobic colony forming units (CFU) per sample. A Taqman-based® real-time PCR protocol was used for quantification of total bacteria, P. gingivalis and A. actinomycetemcomitans. The Wilcoxon signed rank test and McNemar tests were used to evauate the statistical significance of differences between the paired samples. Accordance between the different microbiological methods (conventional PCR versus real-time and PCR and culture versus real-time PCR) was tested using Kappa and Spearman‘s correlation tests. Association between the studied pathogens was evaluated using the chisquare test. P-values < 0.05 were considered statistically significant. Results: Both cultivation and real-time PCR showed significantly (p=0.048 and p=0.0154) higher numbers of bacteria in subgingival plaque samples from combined lesions than in those from simple periodontal lesions. After real-time quantification, the proportions of P. gingivalis and A. actinomycetemcomitans in total number of bacteria were calculated. No significant difference was demonstrated between combined and simple lesions. Conventional PCR demonstrated 57.9%, 21.1% and 42.86% positive samples for A. actinomycetemcomitans 78.9%, 42.1% and 78.6% for P. gingivalis, 42.11%, 21.1% and 35.7% for P. intermedia, 68.4%, 21.1% and 64.3% for T. forsythensis, and 47.4%, 26.3% and 35.7% positive samples for T. denticola in the three types of samples, respectively. Periodontal samples from combined periodontal-endodontic lesions demonstrated significant associations between P. gingivalis and T. denticola (p=0.033) and between P. gingivalis and T. forsythensis (p=0.035). From all the samples that tested positive for P. gingivalis, heteroduplexes analysis revealed that 18 of these contained one strain, 14 showed two strains and only two had three different strains. Conclusions: The significantly higher bacterial levels in subgingival plaque samples from combined periodontal lesions indicate bacterial migration between the root canal and the periodontal pocket. The positive correlation between the bacterial levels of subgingival and endodontic samples from combined periodontal-endodontic lesions also indicates a bacterial communication between the two compartments. The presence of the same strain(s) in periodontal and endodontic samples from combined lesions support the idea that such lesions may represent a single pathological entity.