Two simple self-report measures have been used to assess the heaviness of smoking,‘number of cigarettes per day’(CPD) and‘time to the first cigarette of the day’(TTF). Little attention, however, has been given to the precise method of scoring this information. Using biochemical indicators of heaviness of smoking (alveolar carbon monoxide and cotinine), we explore the optimum data transformations for regression analysis and categorical analysis. We suggest a four category scoring scheme for both time to the first cigarette of the day (leqslant R: less-than-or-eq, slant5, 6–30, 31–60 and 61 + mm) and average daily consumption of cigarettes (1–10, 11–20, 21–30, 31 + cigarettes) as the most powerful and practical categorical scoring of these variables. Due to possible ceiling effects on biochemical measures, we suggest using logarithmic transformations of CPD or TTF for regression or correlation analyses.