ImportanceWhile smoking is associated with a decreased incidence of cutaneous melanoma, the association of smoking with melanoma progression and death is not well defined.ObjectiveTo determine the association of smoking with survival in patients with early-stage primary cutaneous melanoma.Design, Setting, and ParticipantsThis cohort study performed a post hoc analysis of data derived from the randomized, multinational first and second Multicenter Selective Lymphadenectomy Trials (MSLT-I and MSLT-II). Participants were accrued for MSLT-I from January 20, 1994, to March 29, 2002; MSLT-II, from December 21, 2004, to March 31, 2014. Median follow-up was 110.0 (IQR, 53.4-120.0) months for MSLT-I and 67.6 (IQR, 25.8-110.2) months for MSLT-II. Patients aged 18 to 75 years with clinical stages I or II melanoma with a Breslow thickness of 1.00 mm or greater or Clark level IV to V and available standard prognostic and smoking data were included. Analyses were performed from October 4, 2022, to March 31, 2023.ExposureCurrent, former, and never smoking.Main Outcomes and MeasuresMelanoma-specific survival of patients with current, former, and never smoking status was assessed for the entire cohort and for nodal observation and among subgroups with sentinel lymph node biopsy (SLNB)–negative and SLNB-positive findings.ResultsOf 6279 included patients, 3635 (57.9%) were men, and mean (SD) age was 52.7 (13.4) years. The most common tumor location was an extremity (2743 [43.7%]), and mean (SD) Breslow thickness was 2.44 (2.06) mm. Smoking status included 1077 (17.2%) current, 1694 (27.0%) former, and 3508 (55.9%) never. Median follow-up was 78.4 (IQR, 30.5-119.6) months. Current smoking was associated with male sex, younger age, trunk site, thicker tumors, tumor ulceration, and SLNB positivity. Current smoking was associated with a greater risk of melanoma-associated death by multivariable analysis for the entire study (hazard ratio [HR], 1.48 [95% CI, 1.26-1.75]; P < .001). Former smoking was not. The increased risk of melanoma-specific mortality associated with current smoking was greatest for patients with SLNB-negative melanoma (HR, 1.85 [95% CI, 1.35-2.52]; P < .001), but also present for patients with SLNB-positive melanoma (HR, 1.29 [95% CI, 1.04-1.59]; P = .02) and nodal observation (HR, 1.68 [95% CI, 1.09-2.61]; P = .02). Smoking at least 20 cigarettes/d doubled the risk of death due to melanoma for patients with SLNB-negative disease (HR, 2.06 [95% CI, 1.36-3.13]; P < .001).Conclusions and RelevanceThe findings of this cohort study suggest that patients with clinical stage I and II melanoma who smoked had a significantly increased risk of death due to melanoma. Smoking status should be assessed at time of melanoma diagnosis and may be considered a risk factor for disease progression.