Retrospective cohort study. To develop and evaluate a new grading system for destructive lumbar spondyloarthropathy (DLSA) by assessing bony destruction of the facet joints; to evaluate interrater reliability; and to determine the association between anteroposterior (AP) dural sac diameter at the lumbar level and the new grading. The characteristics of DLSA are unknown, hindering clinical care and research. Imaging to determine the cause of DLSA may positively contribute to patient outcome or well-being by providing prognostic information. In the magnetic resonance images (MRIs), we measured an axial midline AP dural sac diameter and evaluated bone destruction caused by amyloidosis at the level of the center of each lumbar disk of the lumbar spine. Two orthopedic surgeons independently evaluated each case at two-month intervals and assigned the grade by rating bone destruction at each lumbar level. Weighted κ and intraclass correlation coefficients for interrater reliability were calculated. In addition, the correlation between the AP diameter of the spinal dural sac at the lumbar level and the new MRI-based DLSA grade was examined. The sample size of 82 patients was reached by examining records of 118 consecutive patients. The mean (SD) age of the included patients was 65 (7.2) years, and 36 (43.9%) were women. The grading of DLSA showed moderate to good interrater reliability at both assessments (κ, 0.59-0.78). Intraclass correlation coefficient showed substantial to excellent agreement (intraclass correlation coefficient, 0.63-0.86). The AP diameter of the spinal dural sac at the lumbar level showed a significant correlation with the new grading ( P <0.001). The new MRI-based grading system for DLSA has good interrater reliability, although the strength of agreement varies somewhat. The new grading system correlates with AP dural sac diameter. Thus, this classification focused on facet erosion, which leads to functional incompetence may be helpful in surgical decision-making.