Objective: Lumbar discography has been widely performed as a diagnostic criterion for discogenic low back pain, but the use of discography is still in a controversy since its advent.The purpose of the current study was to determine when the patients complained of probable symptoms of discogenic low back pain, had obvious disc pathological changes on MRI but showed negative in discography whether we could absolutely exclude the diagnosis of discogenic pain or not.Methods: The patients we selected in our study had only one segmental disc pathology on MRI for the purpose of minimizing the interference and all patients complained of probable diseogenic pain symptoms.Some special lumbar diseases were excluded by physical examination and CT or MRIscan.Finally, 23 isolated patients with negative discography were enrolled in our study from January 2006 to June 2009.12 patients were randomly distributed in experimental group, who underwent provocative discography plus intradiscal injection of Dexamethasone (5mg in each patient), while the other 11 patients in the control group received intradiscal injection of saline as a placebo.Visual analog scale and Oswestry Disability Index were used to evaluate the changes of pain and function after discography.Results: The mean age of the 10 male and 13 female patients was 45.3 years, whose main clinical manifestations included axial back pain (100%), pain in the region of Groin (33.3%), pain in the anterior or posterior region of thigh (42.2%), buttock pain (24.4%) and lower extremity pain (11.1%), and among them L3-L4 pathological disc was found in 2 patient, L4-L5 pathological disc in 13 patients and L5-S1 pathological disc in 8 patients.Discography and CT scans of the 23 patients revealed that there were grade 5 in 19 discs and grade 4 in 4 discs by means of Modified Dallas Classification.After about 6 months follow-up, the results showed that VAS scores and ODI scores were greatly improved during the first 3-month follow-up in the experimental group compared with pre-discography and between groups (p<0.05), but no statistically significant difference after 24-week follow-up (p>0.05).The probable symptoms of discogenic low back pain were significantly improved in the experimental group in the short term.Conclusions: The study indicated that negative discography in patients with probable symptoms of discogenic low back pain can not absolutely exclude the diagnosis of discogenic pain.Patients of this kind may suffer from other diseases, but we can not ignore the existence of discogenic pain.