Abnormal spine biomechanics are associated with various orthopedic disorders. Identifying key biomechanical factors predictive of spinal abnormalities can improve diagnosis and treatment. This study aimed to determine whether specific pelvic biomechanical parameters are significant predictors of spinal abnormalities. We hypothesized that patients with abnormal spine conditions exhibit distinct pelvic measurements compared to those with normal spine conditions. A retrospective analysis was conducted on 1,181 patient records from January to March 2024, focusing on pelvic incidence (PI), pelvic tilt (PT), lumbar lordosis (LL), sacral slope (SS), pelvic radius (PR), and spondylolisthesis. Data were collected from a centralized orthopedic patient database, which integrates de-identified records from the author’s institution and affiliated facilities under the Ministry of Health. This ensured a standardized approach to data entry, with regular audits to maintain accuracy and reliability. Patients’ spine conditions were classified as normal or abnormal based on imaging and clinical examinations. Descriptive statistics summarized the data, and comparative analyses were performed to differentiate between normal and abnormal groups. Decision trees and logistic regression were used to identify significant predictors of spinal abnormalities. Model validation was performed using ROC analysis and 10-fold cross-validation. Preliminary analysis found significant differences between normal and abnormal groups for various factors. Logistic regression identified pelvic incidence, lumbar lordosis, sacral slope, and pelvic radius as significant predictors (P < 0.05). Decision trees classified 69.5% of cases accurately based on pelvic incidence thresholds. Models were validated using ROC analysis (AUC > 0.7) and 10-fold cross-validation (accuracy > 60%). This study provides valuable insights into spine biomechanics by identifying key predictors of spinal abnormalities, particularly pelvic incidence. The decision tree and logistic regression models demonstrated strong predictive capabilities. While prior studies have identified correlations between pelvic parameters and spinal disorders, this research quantifies these associations through predictive modeling, offering practical applications for early diagnosis and intervention. These findings offer the potential for improved diagnostic and treatment strategies for spine disorders. Further prospective studies are necessary to validate these results and enhance predictive models.