It has been well established that both meteorological attributes and air pollution concentrations affect human health outcomes. We examined all cause nonaccident mortality relationships for 28 years (1981-2008) in relation to air pollution and synoptic weather type (encompassing air mass) data in 12 Canadian cities. This study first determines the likelihood of summertime extreme air pollution events within weather types using spatial synoptic classification. Second, it examines the modifying effect of weather types on the relative risk of mortality (RR) due to daily concentrations of air pollution (nitrogen dioxide, ozone, sulfur dioxide, and particulate matter <2.5 μm). We assess both single- and two-pollutant interactions to determine dependent and independent pollutant effects using the relatively new time series technique of distributed lag nonlinear modeling (DLNM). Results display dry tropical (DT) and moist tropical plus (MT+) weathers to result in a fourfold and twofold increased likelihood, respectively, of an extreme pollution event (top 5 % of pollution concentrations throughout the 28 years) occurring. We also demonstrate statistically significant effects of single-pollutant exposure on mortality (p < 0.05) to be dependent on summer weather type, where stronger results occur in dry moderate (fair weather) and DT or MT+ weather types. The overall average single-effect RR increases due to pollutant exposure within DT and MT+ weather types are 14.9 and 11.9 %, respectively. Adjusted exposures (two-way pollutant effect estimates) generally results in decreased RR estimates, indicating that the pollutants are not independent. Adjusting for ozone significantly lowers 67 % of the single-pollutant RR estimates and reduces model variability, which demonstrates that ozone significantly controls a portion of the mortality signal from the model. Our findings demonstrate the mortality risks of air pollution exposure to differ by weather type, with increased accuracy obtained when accounting for interactive effects through adjustment for dependent pollutants using a DLNM.