In recent years, there has been an increasing interest in local infiltration analgesia (LIA) as a technique to control postoperative pain. We conducted a systematic review of randomized clinical trials investigating LIA for total knee arthroplasty (TKA) and total hip arthroplasty (THA) to evaluate the analgesic efficacy of LIA for early postoperative pain treatment. In addition, the analgesic efficacy of wound catheters and implications for length of hospital stay (LOS) were evaluated. Twenty-seven randomized controlled trials in 756 patients operated on with THA and 888 patients operated on with TKA were selected for inclusion in the review. In THA, no additional analgesic effect of LIA compared with placebo was reported in trials with low risk of bias when a multimodal analgesic regimen was administered perioperatively. Compared with intrathecal morphine and epidural analgesia, LIA was reported to have similar or improved analgesic efficacy. In TKA, most trials reported reduced pain and reduced opioid requirements with LIA compared with a control group treated with placebo/no injection. Compared with femoral nerve block, epidural or intrathecal morphine LIA provided similar or improved analgesia in the early postoperative period but most trials had a high risk of bias due to different systemic analgesia between groups. Overall, the use of wound catheters for postoperative administration of local anaesthetic was not supported in the included trials, and LOS was not related to analgesic efficacy. Despite the many studies of LIA, final interpretation is hindered by methodological insufficiencies in most studies, especially because of differences in use of systemic analgesia between groups. However, LIA provides effective analgesia in the initial postoperative period after TKA in most randomized clinical trials even when combined with multimodal systemic analgesia. In contrast, LIA may have limited additional analgesic efficacy in THA when combined with a multimodal analgesic regimen. Postoperative administration of local anaesthetic in wound catheters did not provide additional analgesia when systemic analgesia was similar and LOS was not related to use of LIA with a fast-track set-up.