Approximately 100 toxicants have been identified in cigarette smoke, to which exposure has been linked to a range of serious diseases in smokers. Smoking machines have been used to quantify toxicant emissions from cigarettes for regulatory reporting. The World Health Organization Study Group on Tobacco Product Regulation has proposed a regulatory scenario to identify median values for toxicants found in commercially available products, which could be used to set mandated limits on smoke emissions. We present an alternative approach, which used quantile regression to estimate reference percentiles to help contextualise the toxicant yields of commercially available products with respect to a reference analyte, such as tar or nicotine. To illustrate this approach we examined four toxicants (acetone, N′-nitrosoanatabine, phenol and pyridine) with respect to tar, and explored International Organization for Standardization (ISO) and Health Canada Intense (HCI) regimes. We compared this approach with other methods for assessing toxicants in cigarette smoke, such as ratios to nicotine or tar, and linear regression. We concluded that the quantile regression approach effectively represented data distributions across toxicants for both ISO and HCI regimes. This method provides robust, transparent and intuitive percentile estimates in relation to any desired reference value within the data space.