With the rapid increase in global warming, the impact of extreme temperatures on morbidity and mortality related to respiratory diseases has attracted considerable attention. In the current study, we quantified the relative risks (RRs) of mortality for respiratory diseases in three capital cities in Northeast China. We used a distributed lag nonlinear model (DLNM) based on a generalized additive model (GAM) to estimate the impact of extreme temperatures on respiratory mortality in Shenyang, Changchun, and Harbin from 2014 to 2016. The results revealed that the maximum cumulative RRs and 95% confidence intervals (CIs) were 1.52 (1.28-1.80), 1.42 (1.07-1.89), and 1.38 (1.21-1.58) in Shenyang, Changchun, and Harbin respectively when the median temperature was used as reference. The effect of extremely high temperature (99th percentile relative to 90th percentile) on respiratory mortality was found to be strongest in Shenyang (at the lowest latitude), while the effect of extreme low temperature (1st percentile relative to 10th percentile) on respiratory mortality was strongest in Harbin (at the highest latitude). In Shenyang and Changchun, the effects of high temperatures were much more intense and pronounced in females. Furthermore, the effect of high temperatures was more acute, whereas the effect of low temperatures was longer lasting.