Cigarette users are more susceptible than non-smokers to periodontitis, a bacterial-induced, inflammation-driven, destructive disease of the supporting tissues of the teeth. We hypothesized that clinical periodontal findings and microbiological and/or inflammatory marker levels would be intermediate in those exposed to environmental tobacco smoke compared to active smokers and non-smokers.Sixty individuals were recruited from a University periodontal clinic and assigned as non-smokers, active smokers or passive-smokers according to their self reports. Clinical periodontal measurements, comprising plaque index, probing depth (PD), clinical attachment level (CAL) and bleeding on probing, were recorded at six sites per tooth. Cotinine levels were determined in whole saliva samples by EIA. and infection was determined by PCR, while matrix metalloproteinase-8 (MMP-8) and interleukin-8 (IL-8) concentrations were determined by ELISA.Study groups were subsequently reassigned in accordance with the cotinine data. The smoker group exhibited higher mean PD and CAL values compared to the non-smoker group (p<0.05). Passive-smokers exhibited PD and CAL values smaller than those of the active smokers and greater than those of the non-smokers, but the differences were not statistically significant. PD and CAL values correlated with cotinine concentrations (p<0.05). infection was noted in most subjects, irrespective of smoking status. infection was noted in 4/23 (17.4%) smokers, 0/16 (0%) environmentally-exposed recruits and 2/21 (9.5%) non-smokers. Salivary MMP-8 and IL-8 levels were lower in smokers compared to both non-smokers and passive-smokers but the differences were not significant (all p>0.05).The present clinical periodontal findings provide further support for a negative, dose-related effect of tobacco exposure on periodontal health. The tendency for a more prevalent detection of and for a suppressed inflammatory response observed in the smokers may partly explain the increased susceptibility to periodontal tissue destruction, but needs to be verified in larger scale studies.