While low back pain (LBP) constitutes a global life disorder cause, the contribution of paraspinal muscles to its pathogenicity remains elusive. We characterized the paraspinal muscles of patients with LBP using lumbar three-dimensional computed tomography (CT) and magnetic resonance imaging (MRI) mDIXON-Quant, and evaluated the risk factors combined with clinical data. Methods: A retrospective study involving 181 patients (10-40 years) who underwent lumbar 3D-CT and MRI mDIXON from January 1, 2021 to December 31, 2022, and divided into normal, non-chronic LBP [non-CLBP], and CLBP groups. Clinical data, paraspinal muscle cross-sectional area, Hounsfield unit for CT values, and fat fraction derived from mDIXON-Quant were compared. Three readers analyzed the images independently; intra-and interobserver agreement was measured. Spearman analysis and multiple logistic regression were used to analyze the correlation between clinical data, radiologic and paraspinal muscle parameters. A nomogram was constructed for individualized prediction. Results: Correlation analysis revealed that body mass index, visual analog scale score, Pfirrmann grade, annulus fibrosus tear, lumbar lordosis (LL), and Modic changes correlated with LBP (all P<0.05). The Pfirrmann grade and annulus fibrosus tear showed positive correlation (r=0.673, 0.559), whereas LL was negatively correlated (r=−0.469). The multifidus CT values were negatively correlated with LBP at L4-5/L5-S1; the multifidus fat fraction was positively correlated at L4-5/L5-S1 (r=0.734, r=0.584, P<0.001). The multiple logistic regression showed that L4-5 multifidus fat fraction (P=0.046, OR=1.167), Pfirrmann grade (P=0.017, OR=0.063), LL (P=0.002, OR=0.828) and annulus fibrosus tear (P=0.005, OR=0.024) were risk factors for predicting LBP in the non-CLBP group; in the CLBP group, BMI (P=0.048 OR=1.225), L4-5 multifidus fat fraction (P=0.001 OR=1.299), LL (P=0.003, OR=0.841) and Pfirrmann classification (P=0.009, OR=0.046) were risk factors. Conclusion: BMI, L4-5 multifidus fat fraction, LL, and Pfirrmann grade are risk factors for CLBP in patients under 40; whereas annulus fibrosus tear is an independent risk factor for non-CLBP, nomograms derived from these parameters can help predict LBP and MRI mDIXON-Quant is recommended for quantitatively analyzing paraspinal muscle fat infiltration.