<p>Deficiency in Vitamin B1, an essential water-soluble micronutrient and coenzyme for key metabolic pathways, has been associated with poor clinical outcomes in critically ill patients with deficiency. Despite its clinical significance, there remains a lack of validated bedside tools for identifying high-risk individuals with vitamin B1 deficiency. This study aimed to develop and validate a nomogram-based prediction model using readily available clinical data upon admission to facilitate early screening and intervention. A prospective cohort study was conducted, enrolling 340 critically ill patients who were randomly allocated into a training cohort (<i>n</i> = 237) and a validation cohort (<i>n</i> = 103) at a 7:3 ratio. Feature selection was performed using LASSO regression, followed by logistic regression modeling, which was subsequently visualized as a nomogram. The model’s predictive performance was assessed based on discrimination, calibration, and clinical applicability. The overall incidence of vitamin B1 deficiency was 16.2%. Eight independent predictors were identified: lymphocyte percentage (LYMPH%), C-reactive protein (CRP), D-dimer, albumin levels, body mass index (BMI), sepsis classification, smoking status, and alcohol consumption history. The nomogram exhibited strong discriminative ability in the training cohort (AUC = 0.802, 95% CI: 0.720–0.885) and moderate performance in the validation cohort (AUC = 0.610, 95% CI: 0.469–0.750). Calibration curves demonstrated excellent agreement between predicted and observed probabilities, while decision curve analysis confirmed its clinical utility. In conclusion, this study successfully established a nomogram-based predictive model for vitamin B1 deficiency in critically ill patients, offering a practical and reliable tool for clinical decision-making.</p>