Abstract: Objective To explore the risk factors of microvascular invasion (MVI) in solitary small hepatocellular carcinoma based on preoperative serological indicators and imaging examination, and to construct a nomogram prediction model. Methods A total of 205 patients with solitary small hepatocellular carcinoma who underwent radical hepatectomy at Department of Hepatobiliary Pancreatic Surgery, the Affiliated Hospital of Xuzhou Medical University, from June 2020 to December 2023 were enrolled, and their clinical data were retrospectively analyzed. The patients were randomly divided into a training group (n=143) and a validation group (n=62) in a 7∶3 ratio. LASSO regression and logistic regression analysis were used to identify risk factors for MVI, and a nomogram prediction model was constructed. Receiver operating characteristic (ROC) curves, calibration curve, and clinical decision curve were plotted to evaluate the model's efficiency, and model performance was verified in the validation group. Results LASSO and logistic regression analyses revealed that tumor maximum diameter>2.65 cm, protein induced by vitamin K absence or antagonist-Ⅱ (PIVKA-Ⅱ)>40 AU/L, neutrophil-to-lymphocyte ratio (NLR)>1.92, fibrinogen-to-albumin ratio (FAR)>0.056, and incomplete capsule were independent risk factors for MVI in patients with solitary small hepatocellular carcinoma. A nomogram prediction model was constructed based on these indicators. ROC curves for predicting MVI in the training and validation groups showed that the combined model in both groups had a higher predictive value than single-variable models, with areas under the curve (AUC) of 0.866 (95%CI 0.799-0.932) for the training group and 0.854 (95%CI 0.748-0.960) for the validation group. The Hosmer-Lemeshow test showed a high degree of fit between predicted probabilities and postoperative pathology results, with χ²=4.57, P=0.802 in the training set, and χ²=11.4, P=0.122 in the validation set, indicating good consistency of the predictive model. Decision curve analysis showed threshold probabilities of 0.05-0.75 for the training group and 0.05-0.65 for the validation group, suggesting that the model has high predictive ability and can effectively predict MVI occurrence. Conclusions The nomogram prediction model constructed based on tumor maximum diameter>2.65 cm, PIVKA-II>40 AU/L, NLR>1.92, FAR>0.056, and incomplete capsule has high value for preoperative prediction of MVI in patients with solitary small hepatocellular carcinoma, providing guidance for clinical decision-making.