In Canada, the Canadian Council of Ministers for the Environment (CCME) is currently engaged in a process to determine how best to reduce air emissions from oil refineries. The National Framework for Petroleum Refineries Emissions Reduction (NFPRER) is being developed with the input of stakeholders, including nongovernment organizations (NGOs), industry, and regulatory jurisdictions. One component of this framework is the development of a tool to prioritize emissions for reduction based on estimated health impacts. HEIDI II (Health Effects Indicators Decision Index II) is a spreadsheet-based model that prioritizes a series of carcinogenic and noncarcinogenic air toxicicants and criteria air contaminants commonly emitted from Canadian oil refineries. A generic meteorological dispersion model was applied to reported annual emissions data for each of Canada's 20 refineries. Photodegradation rates and ambient levels of each substance were accounted for, and air concentrations were calculated for 20 geographic zones around each refinery. These were coupled to toxicity data derived mainly from Health Canada and the U.S. Environmental Protection Agency (EPA), and applied to target populations of children, adults and seniors. HEIDI II predicts incidence of relevant disease endpoints from each substance emitted, except for benzene, toluene, ethylbenzene, and xylene (BTEX) and polycyclic aromatic hydrocarbons (PAH), which were treated as chemical mixtures. Rankings were based on predicted case incidence or the application of a common health impact metric, disability-adjusted life years (DALYs), to the predicted incidence. Using the DALY approach, priority rankings can be made within each of the chemical classes, or across all three classes together. HEIDI II incorporates several switches that allow the user to investigate alternate scenarios based on stack height, average daily sunlight hours (for calculating photodegradation), and the possibility of emissions below regulatory reporting thresholds.