BACKGROUND: Major knee surgery is a common operative procedure to help people with end-stage knee disease or trauma to regain mobility and have improved quality of life. Poorly controlled pain immediately after surgery is still a key issue for this procedure. Peripheral nerve blocks are localized and site-specific analgesic options for major knee surgery. The increasing use of peripheral nerve blocks following major knee surgery requires the synthesis of evidence to evaluate its effectiveness and safety, when compared with systemic, local infiltration, epidural and spinal analgesia. OBJECTIVES: To examine the efficacy and safety of peripheral nerve blocks for postoperative pain control following major knee surgery using methods that permit comparison with systemic, local infiltration, epidural and spinal analgesia. SEARCH METHODS: We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 1, 2014), MEDLINE and EMBASE, from their inception to February 2014. We identified ongoing studies by searching trial registries, including the metaRegister of controlled trials (mRCT), clinicaltrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP). SELECTION CRITERIA: We included participant-blind, randomized controlled trials of adult participants (15 years or older) undergoing major knee surgery, in which peripheral nerve blocks were compared to systemic, local infiltration, epidural and spinal analgesia for postoperative pain relief. DATA COLLECTION AND ANALYSIS: Two review authors independently assessed study eligibility and extracted data. We recorded information on participants, methods, interventions, outcomes (pain intensity, additional analgesic consumption, adverse events, knee range of motion, length of hospital stay, hospital costs, and participant satisfaction). We used the 5-point Oxford quality and validity scale to assess methodological quality, as well as criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions. We conducted meta-analysis of two or more studies with sufficient data to investigate the same outcome. We used the I² statistic to explore the heterogeneity. If there was no significant heterogeneity (I² value 0% to 40%), we used a fixed-effect model for meta-analysis, but otherwise we used a random-effects model. For dichotomous data, we present results as a summary risk ratio (RR) and a 95% confidence interval (95% CI). Where possible, we calculated the number needed to treat for an additional beneficial outcome (NNTB) or for an additional harmful outcome (NNTH), together with 95% CIs. For continuous data, we used the mean difference (MD) and 95% CI for similar outcome measures. We describe the findings of individual studies where pooling of data was not possible. MAIN RESULTS: According to the eligibility criteria, we include 23 studies with 1571 participants, with high methodological quality overall. The studies compared peripheral nerve blocks adjunctive to systemic analgesia with systemic analgesia alone (19 studies), peripheral nerve blocks with local infiltration (three studies), and peripheral nerve blocks with epidural analgesia (one study). No study compared peripheral nerve blocks with spinal analgesia.Compared with systemic analgesia alone, peripheral nerve blocks adjunctive to systemic analgesia resulted in a significantly lower pain intensity score at rest, using a 100 mm visual analogue scale, at all time periods within 72 hours postoperatively, including the zero to 23 hours interval (MD -11.85, 95% CI -20.45 to -3.25, seven studies, 390 participants), the 24 to 47 hours interval (MD -12.92, 95% CI -19.82 to -6.02, six studies, 320 participants) and the 48 to 72 hours interval (MD -9.72, 95% CI -16.75 to -2.70, four studies, 210 participants). Subgroup analyses suggested that the high levels of statistical variation in our analyses could be explained by larger effects in people undergoing total knee arthroplasty compared with other types of surgery. Pain intensity was also significantly reduced on movement in the 48 to 72 hours interval postoperatively (MD -6.19, 95% CI -11.76 to -0.62, two studies, 112 participants). There was no significant difference on movement between these two groups in the time period of zero to 23 hours (MD -6.95, 95% CI -15.92 to 2.01, five studies, 304 participants) and 24 to 47 hours (MD -8.87, 95% CI -27.77 to 10.03, three studies, 182 participants). The included studies reported diverse types of adverse events, and we did not conduct a meta-analysis on specific types of adverse event. The numbers of studies and participants were also too few to draw conclusions on the other prespecified outcomes of: additional analgesic consumption; median time to remedication; knee range of motion; median time to ambulation; length of hospital stay; hospital costs; and participant satisfaction. There were insufficient data to compare peripheral nerve blocks and local infiltration or between peripheral nerve blocks and epidural analgesia. AUTHORS' CONCLUSIONS: All of the included studies reported the main outcome of pain intensity but did not cover all the secondary outcomes of interest. The current review provides evidence that the use of peripheral nerve blocks as adjunctive techniques to systemic analgesia reduced pain intensity when compared with systemic analgesia alone after major knee surgery. There were too few data to draw conclusions on other outcomes of interest. More trials are needed to demonstrate a significant difference when compared with local infiltration, epidural analgesia and spinal analgesia.