Gingival bleeding is an objective, easily assessed sign of inflammation that is associated with several periodontal diseases. Many bleeding indices have been devised; some assess bleeding as simply present or absent, whereas others use grading in an attempt to assess severity of bleeding. The choice of which index to use depends on whether the purpose is an epidemiological survey, a clinical study, diagnosis and treatment, or patient motivation. Bleeding may be elicited manually with toothpicks, dental floss or a periodontal probe, but a controlled-force probe, although more expensive, causes less trauma and less false-positive bleeding from healthy tissues. As a predictor of periodontal disease progression, bleeding on probing has low sensitivity owing to a high frequency of false-positive responses, but has high specificity in that failure to bleed indicates health. There is evidence that smokers have less, or delayed, gingival bleeding when compared with non-smokers; therefore smoking needs to be controlled for in studies of gingival bleeding. Measurement of gingival bleeding tendency should be an integral part of a comprehensive oral examination. In clinical practice, the use of a graded bleeding index is more likely to identify sites that are at risk of further destructive activity. For monitoring individual patients, both for response to initial therapy and during maintenance, a modified Sulcus Bleeding Index (mSBI) with three bleeding scores is recommended in preference to dichotomous scoring of bleeding.