We report our experience of treating lumbar herniated disc by intradiscal injection of an oxygen-ozone mixture. Ozone (03, MW = 48) is a triatomic molecule, having antiviral, disinfectant and antiseptic properties. Several mechanisms of action have been proposed to explain the efficacy of the treatment: 1) analgesic action; 2) anti-inflammatory action; 3) oxidant action on the proteoglycan in the nucleus pulposus. We treated 93 patients (50 women, 43 men) aged from 24 to 45 yrs (average age 38 yrs) from June 1996 to April 1998. All patients presented sciatica and/or low back pain, lasting two or more months; patients had in the mean time received both medical and physical therapy with mild or no benefit. Diagnostic tests in all patients included plain film x-ray, CT and/or MR at the level of the lumbar spine disclasing a herniated or protruded disc with nerve root or thecal sac compression. We divided patients to be treated in to two groups: the first one group included 35 patients already selected for surgery who presented herniated or protruded disc with radicular pain with associated neurological deficit (hypoesthesia and partial loss of reflex). Those patients had already had medical and physical therapy for two or more months and agreed to try the percutaneous treatment before surgery. CT or MR in this group demonstrated the presence of intraforaminal, extra or sub-ligamentary and sequestrated herniated disc. The second group included 58 patients with radicular pain but without neurological deficit; patients in this group had received medical and/or physical therapy for two or more months and CT showed the presence of a small subligamentary herniated or protruded disc. We considered the results according to the modified MacNab method. In the first group we had “failure” in all patients; in seven cases the symptoms improved for one month, but recurred later on. In the second group 45 patients had “success” showing complete clinical recovery within five to six days after treatment, all remained without symptoms up to six months or more of follow-up. The remaining 13 patients presented the same symptoms again within three months after a temporary clinical recovery. The goal of this study was to present this new technique that can also be compared with a previous study of different percutaneous treatment. Clinical and neuroradiological indications and the contraindications are well known, and must be followed to achieve good results and avoid complications.