OBJECTIVE: Multiple-level Intervertebral disc degeneration (IDD) in patients with lumbar disc herniation (LDH) is related to postoperative re-herniation and low back pain. Although many investigators believed that there is an interdependence between paraspinal muscles degeneration and IDD, few studies focused on the fatty infiltration of paraspinal muscles on single- and multiple-level IDD in patients with LDH. This study aims to investigate the difference on the fatty infiltration of paraspinal muscles between single- and multiple-levels IDD in patients with LDH. and to explore in patients with LDH whether fatty infiltration is a potential risk factor for multiple-level IDD. METHODS: This study was conducted as a retrospective observational analysis of 82 patients with LDH from January 1, 2020 to December 30, 2020 in our hospital were enrolled. Twenty-seven cases had single-level IDD (Group A), and 55 cases had multiple-level IDD (Group B). We measured the mean computed tomography (CT) density value of the paraspinal muscles, including multifidus (MF), erector spinae (ES) and psoas muscle (PM) at each disc from L1 to S1. Subgroups were set to further analyze the odds ratio (OR) of fatty infiltration of paraspinal muscles in different sex and BMI groups. We measured sagittal angles and analyzed the relationships between these angles and IDD. Finally, we use logistic regression, adjusted for other confounding factors, to investigate whether fatty infiltration is an independent risk factor for multi-level IDD. RESULTS: The average age in multi-level IDD (51.40 ± 15.47 years) was significantly higher than single-level IDD (33.37 ± 7.10 years). The mean CT density value of MF, ES and PM in single-level IDD was significantly higher than multi-level IDD (all ps < 0.001). There was no significant difference of the mean value of angles between the two groups. No matter being fat (body mass index [BMI] > 24.0 kg/m) or normal, patients with low mean muscle CT density value of MF and ES are significantly easier to suffer from multiple-level IDD. In the pure model, the average CT density value of the MF, ES and PM is all significantly associated with the occurrence of multi-IDD. However, after adjusting for various confounding factors, only the OR of the average CT density value for MF and ES remains statistically significant (OR = 0.810, 0.834, respectively). CONCLUSIONS: In patients with LDH, patients with multiple-level IDD have more severe fatty infiltration of MF and ES than those with single-level IDD. Fatty infiltration of MF and ES are independent risk factors for multiple-level IDD in LDH patients.