Abstract The nature of symptoms and signs of sciatica, genetic factors, and efficacy of periradicular infiltration were studied in 160 nonoperated patients with unilateral sciatica of 3 to 28 weeks duration. Back and leg pain (100-mm VAS), disability (Oswestry), and quality-of-life (NHP) were evaluated. ENMG and 1.5-T MRI were performed on every patient. Presence of the Trp2 and Trp3 alleles of collagen IX was determined from blood samples. After informed consent, patients were randomized for periradicular infiltration with either methylprednisolone-bupivacaine, or saline. The final follow-up assessment was 1 year after the intervention. Economic analysis was based on data gathered from the patients, medical records and the National Insurance Register. At baseline, symptoms of sciatica did not correlate with the type of displacement of the symptomatic disc in MRI, or the presence of the Trp2 or Trp3 alleles. In the case of the Trp2 allele, there was a non-significant tendency for the presence of a radial tear at the L4–5 level. A significant genotype-phenotype association was found for the Trp3 allele: 15 of 34 (44%) patients with the Trp3 allele were positive for thoracolumbar Scheuermann's disease in MRI compared to 19% for sciatic patients without the allele (p = 0.003). Periradicular infiltration with methylprednisolone-bupivacaine produced a significant treatment effect compared to saline at 2 weeks for leg pain, straight leg raising, lumbar flexion and patient satisfaction. At 6 months, saline was superior to steroid in back and leg pain. By 1 year, 18 patients in the methylprednisolone group and 15 in the saline group had received surgical treatment. Subgroup analysis revealed that the short-term effect of the steroid treatment was most pronounced for contained herniations and symptomatic lesions situated at the L4–5 (or L3–4) disc level. Patients with a contained herniation were less likely to undergo back surgery when receiving the steroid treatment and they also had significantly fewer days on sick leave from 3 to 6 months. Counter-effectiveness was most pronounced for extrusions. The results indicate that disability among sciatic patients may be present even when MRI findings are minor; and vice versa, prominent MRI findings may not associate with any symptoms. However, MRI seems to be useful for identifying patients with the Trp3 allele. On the basis of the treatment intervention results, periradicular infiltration with a combination of steroid and anaesthetic may be recommended for sciatica as it offers at least short-term pain relief. Furthermore, in the case of contained herniations the steroid injection is cost-effective and may also prevent surgery. However, this subgroup analysis calls for a verification study.