Objective To establish and validate a predictive model for non-curative resection of rectal neuroendocrine tumors (R-NETs) ≤20 mm. Methods Data from patients with R-NETs treated at the First Affiliated Hospital of Soochow University and the Suzhou Ninth People''s Hospital from January 2013 to December 2023 were retrospectively analyzed. Clinical data, endoscopic findings, and pathological characteristics were analyzed. Univariate analysis was performed using independent sample t-tests and Chi-square tests. Variables were screened using the forward stepwise binary logistic regression to establish a risk prediction model for non-curative resection of R-NETs ≤20 mm, with subsequent construction of a nomogram. The performance of the model was evaluated using the receiver operating characteristic (ROC) curve. The consistency between predicted and observed probabilities was assessed using calibration curves, and the clinical net benefit of the model was evaluated using decision curve analysis. Results A total of 213 patients were included, with age of 50.53±11.42 years, and 102 (47.9%) were male. The distance of the lesion from the dentate line was 7.11±2.79 cm, and the tumor long diameter was 8.24±3.75 mm. Compared to curative resection cases, non-curative resection cases were more likely to exhibit tumor surface depression, higher tumor G-stage, higher Ki-67 index and higher chromograninA (CgA) positivity rate (P<0.05). Through forward variable selection in binary logistic regression, a model was established with Ki-67 index (P=0.014, OR=1.214, 95%CI: 1.039-1.417), surface depression (P=0.027, OR=2.348, 95%CI: 1.100-5.013), and CgA positivity (P<0.001, OR=5.399, 95%CI: 2.764-10.544) as parameters, with a corresponding nomogram. The area under the ROC curve of the model was 0.766 (95%CI: 0.696-0.837), and clinical decision curve analysis confirmed its good clinical net benefit. The calibration curve showed good consistency between predicted and observed probabilities. Conclusion This study establishes a risk prediction model for non-curative resection of R-NETs ≤20 mm based on surface depression, Ki-67 index, and CgA positivity. The model demonstrates strong predictive performance and offers valuable guidance for clinical endoscopists.