Purpose <p>To compare overall survival (OS) between neoadjuvant chemotherapy (NACT) and adjuvant chemotherapy (ACT) in stage III rectal cancer with tumor deposits (TD), and to develop and validate an ACT-based nomogram for individualized prognostic risk stratification.</p> Methods <p>The study included 3060 stage III rectal cancer patients with tumor deposits from the SEER database (2010–2017). Propensity score matching (1:1 PSM) addressed treatment bias. The adjuvant chemotherapy cohort (<i>n</i> = 1520) was divided into training (<i>n</i> = 1,064) and validation (<i>n</i> = 456) sets at a 7:3 ratio. Multivariable Cox regression analysis identified prognostic factors for nomogram construction. Model performance was assessed via C-index, AUC, and calibration curves. Patients were classified as high-risk and low-risk groups based on the nomogram scores. KM curves were used to compare the survival differences between the two patient categories.</p> Results <p>After 1:1 PSM (497 vs. 497), ACT showed significantly better OS than NACT (log-rank <i>P</i> = 0.0027). In the ACT cohort (<i>n</i> = 1520), a nomogram integrating age, CEA, PNI, T stage, and N stage achieved AUCs of 0.711/0.713/0.716 at 1/3/5 years in the training set and 0.694/0.724/0.698 in the validation set. Using an optimal cut-off of 134.55, patients were stratified into low- and high-risk groups with significantly different OS (<i>P</i> &lt; 0.0001).</p> Conclusion <p>The nomogram was constructed to successfully predict the 1-, 3- and 5-year OS for patients who received adjuvant chemotherapy. In addition, this nomogram may assist in developing clinical treatment strategies for patients with tumor deposits who are undergoing treatment.</p>