BACKGROUND: Standard of care for resectable pancreatic cancer is a combination of surgical resection (SR) and multiagent chemotherapy (MCT). We aim to determine whether SR or MCT is associated with superior survival for patients receiving only single-modality therapy. METHODS: Patients with stage I-IIb pancreatic head adenocarcinoma who received either MCT or SR were identified in the NCDB (2013-2015). Following a piecewise approach to estimating hazards over the course of follow-up, conditional overall survival (OS) at 30, 60, and 90 days after treatment initiation was estimated using landmark analyses. RESULTS: 3103 patients received MCT alone (60.3%) and 2043 underwent SR alone (39.7%). SR had an OS disadvantage at 30 (HR 3.99, 95% CI 3.12-5.11) and 60 days (HR 1.85, 95% CI 1.4-2.45), but an OS advantage after 90 days (HR 0.59, 95% CI 0.55-0.64). In a landmark analysis conditioned on 90 days survival post treatment initiation, median OS was improved for SR (17.0 vs. 12.2 months, p < 0.0001); SR improved 3-year OS by 21.3% (p < 0.05), despite patients being older (median 72 vs. 67 years, p < 0.0001) with higher Charlson-Deyo comorbidity scores (≥2: 11.2 vs. 8.6%, p = 0.006). CONCLUSION: For patients with resectable pancreatic cancer, SR is associated with superior long-term survival compared to MCT.