Following a recent claim that the use of cotinine in body fluids, to assess passive smoking and smoking "deception", was confounded by metabolic individuality, and by non-tobacco sources of dietary nicotine, particularly tea, data were examined from a large cross-sectional survey in a tea-drinking population. In 3383 men and women aged 40-59 years from the Scottish Heart Health Study, defined as non-smokers, both by self-report and by low thiocyanate and expired air carbon monoxide levels, serum cotinine showed minimal association with self-reported daily average tea consumption. However, there was a strong correlation between degree of self-reported passive tobacco smoke exposure and median serum cotinine level. In the same survey, serum cotinine in 4144 self-reported non-smokers and in 3326 smokers showed entirely different distributions, but the same range, suggesting heavy nicotine intake in some "non-smokers". These analyses confirm that cotinine levels in true non-smokers reflect far more the nicotine in inhaled ambient tobacco smoke than they do nicotine in tea. Some smoking "deceivers" have the same degree of exposure to nicotine as heavy smokers. Despite individual variability, the claim of confounding is poorly supported, and cotinine is confirmed as an indicator both of passive smoking and of smoking deception.