Several studies have examined the relationship between smoking and mortality in cancer survivors. However, few have reported the relationships in several cancer sites (i.e., bladder, non-melanoma skin, uterine, melanoma, and lymphoma), and limited data exist on the dose-response relationship between number of cigarettes smoked per day or duration of smoking cessation and mortality. Cancer survivors (N = 35,093, 61% female, mean age = 47 years old) from the National Health Interview Survey with linked data retrieved from the National Death Index served as our study participants. Cox proportional-hazards models were used to assess associations between smoking status, all-cause, and disease-specific mortality. After a median follow-up of 13 years, 11,066 deaths occurred. Survivors who reported smoking at study entry had a 73%, 75%, 85% higher risk for cardiovascular disease, cancer, and all-cause mortality, respectively when compared to nonsmokers. Former smokers had a 31% and 37% higher risk of all-cause and cancer mortality, respectively when compared to nonsmokers. The observed relationships appeared to differ by the number of cigarettes smoked (i.e., ≥ 10 per day), especially for breast, cervix, lung, prostate, uterine and non-melanoma skin cancer survivors. Those who continued smoking post diagnosis were at greatest risk of all-cause and cancer mortality, but the associations varied substantially by cancer site. These data provide sufficient evidence of the health hazards associated with smoking for cancer survivors and provide further support for public health strategies designed to curb smoking in this vulnerable population. AJCR Copyright © 2019.