The effectiveness of surgery in patients with sciatica due to lumbar disc herniations is not without dispute. The goal of this study was to assess the effects of surgery versus conservative therapy (including epidural injections) for patients with sciatica due to lumbar disc herniation. A comprehensive search was conducted in MEDLINE, EMBASE, CINAHL, CENTRAL, and PEDro up to October 2009. Randomised controlled trials of adults with lumbar radicular pain, which evaluated at least one clinically relevant outcome measure (pain, functional status, perceived recovery, lost days of work) were included. Two authors assessed risk of bias according to Cochrane criteria and extracted the data. In total, five studies were identified, two of which with a low risk of bias. One study compared early surgery with prolonged conservative care followed by surgery if needed; three studies compared surgery with usual conservative care, and one study compared surgery with epidural injections. Data were not pooled because of clinical heterogeneity and poor reporting of data. One large low-risk-of-bias trial demonstrated that early surgery in patients with 6–12 weeks of radicular pain leads to faster pain relief when compared with prolonged conservative treatment, but there were no differences after 1 and 2 years. Another large low-risk-of-bias trial between surgery and usual conservative care found no statistically significant differences on any of the primary outcome measures after 1 and 2 years. Future studies should evaluate who benefits more from surgery and who from conservative care.