Objective To analyze the cl inical effects of modified transforaminal lumbar interbody fusion (TLIF) for the treatment of lumbar degenerative disease. Methods From October 2003 to December 2006, 33 patients with lumbar degenerative disease (L3-S1) were treated by modified TLIF. There were 14 males and 19 females with an average age of 52.2 years(33 to 70 years). The median disease course was 1.8 years (4 months to 15 years). A total of 42 levels were fused, including 24 cases of single level and 9 cases of double levels. The results of preoperative diagnosis were lumbar degenerative spondylol isthesis with stenosis (8 cases), isthmic spondylol isthesis (5 cases), degenerative lumbar stenosis (16 cases), huge herniated disc with segmental instabil ity (3 cases) and failed back surgery syndrome (1 case). During the modified TLIF procedure, total inferior facet process and inner half summit of superior facet process of TLIF side were resected to make the posterior wall of foramen opened partly. After the bone graft (3 to 5 mL) was placed into the interbody space, a single rectangle Cage was inserted obl iquely from 30° to 40° toward the midl ine. Combined with pedicle screw instrumentation, TLIF was accompl ished. Middle canal and opposite side nerve root decompression were performed simultaneously when necessary. Results Intraoperative dura mater rupture, postoperative cerebral spinal fluid leakage, deep wound infection and transient nerve root stimulation occurredin 1 case respectively, and were all recovered after treatment. No patients had permanent neurologic deficit or aggravation. All patients were followed up for 20 to 58 months (mean 27.2 months). At the follow-up after 1 year postoperatively, all the operated segments achieved fusion standard and no broken screw or Cage dislocation occurred. All 13 cases of spondylol isthesis were reduced thoroughly and maintained satisfactorily. Nineteen patients remained sl ight chronic back pain. There was significant difference (P lt; 0.05) in JOA score between preoperation (14.9 ± 5.1) and postoperation (25.9 ± 3.0). The rate of cl inical improvement was 80.5% (excellent in 24 cases, good in 7 cases, and fair in 2 cases). Conclusion The modified TLIF carries out the less invasive principles in opening operations, simpl ifies the manipulation and expands the indication of TLIF to some extent, and the cl inical results for the treatment of lumbar degenerative disease is satisfactory.