In both cross-sectional and longitudinal studies of young adults with plaque-induced gingivitis it has been observed that bleeding upon probing is only weakly associated with supragingival plaque. It has been speculated that gingival bleeding may be influenced by several independent factors other than plaque. Great intra- and interindividual variation of gingival thickness and width has been reported. Based on respective observations, the existence of different gingival phenotypes has been suggested. The aim of the present study was to investigate the possible influence of gingival thickness and width on bleeding on probing. Forty young adults with mild, plaque-induced gingivitis, 24 non-smokers and 16 smokers, participated in this cross-sectional study. In addition to periodontal probing depth, clinical attachment loss, width of gingiva, bleeding on probing, and presence of plaque, gingival thickness was measured with an ultrasonic device. Multivariable models were separately calculated for buccal, mandibular lingual, and palatal surfaces and generally adjusted for tooth type. Generalised Estimation Equation methodology was employed in order to adjust for correlated observations. Plaque was significantly associated with bleeding upon probing only at buccal sites (odds ratio 1.80, 95% confidence interval 1.19-2.72) An influence of similar magnitude was identified for smoking (odds ratio 1.76; 1.07-2.89). At lingual sites in the mandible, bleeding was influenced by smoking (odds ratio 2.25; 1.18-4.25) and gingival thickness (odds ratio for thick gingiva >1 mm of 1.93; 1.02-3.65), but not plaque. At palatal sites, only periodontal probing depth had an influence (odds ratio 1.89; 1.25-2.84). It was concluded that, apart from supragingival plaque, smoking was an independent risk factor for gingival bleeding on probing. Thin and vulnerable gingiva of insufficient width was not more likely to bleed after probing than thicker tissue.