Epidemiological studies into the alleged health effects of ETS typically use questionnaires which, at best, quantify ETS ‘exposure’. However, exposure is not, in itself, sufficient; what is required is a measure of the substance that enters the body and reaches the site of action i.e. the dose of the substance. As ETS is a complex mixture various approaches based on components of ETS have been employed to quantify dose. For example, the measurement of cotinine in body fluids has been used but such techniques are also fraught with error. It is recognised that a measure of dose should ideally satisfy four criteria: (1) quantifiable-it should be possible to assess the magnitude of the dose in an objective and accessible way, (2) relevance-it should encompass both concentration and duration, (3) specificity-it should be unique to the substance under consideration, and (4) connection to the end point of concern- the dose measured should be the amount that participates in the potential reaction allegedly associated with the "exposure". Techniques will be appraised in relation to the criteria above and their scope and limitations will be discussed.