This review aimed to identify the efficacy of low-level laser therapy (LLLT) in the management of orthodontic pain. This systematic review and meta-analysis was carried out in accordance with Cochrane Handbook and the PRISMA statement. An extensive literature search for RCTs, quasi-RCTs, and CCTs was performed through CENTRAL, PubMed, Embase, Medline, CNKI, and CBM up to October 2011. Risk of bias assessment was performed via referring to the Cochrane tool for risk of bias assessment. Meta-analysis was implemented using Review Manager 5.1. As a result, four RCTs, two quasi-RCTs, and two CCTs were selected from 152 relevant studies, including 641 patients from six countries. The meta-analysis demonstrated that 24 % risk of incidence of pain was reduced by LLLT (RR = 0.76, 95 % CI range 0.63–0.92, P = 0.006). In addition, compared to the control group, LLLT brought forward “the most painful day” (MD = −0.42, 95 % CI range −0.74– − 0.10, P = 0.009). Furthermore, the LLLT group also implied a trend of earlier end of pain compared with the control group (MD = −1.37, 95 % CI range −3.37–0.64, P = 0.18) and the pseudo-laser group (MD = −1.04, 95 % CI range −4.22–2.15, P = 0.52). However, because of the methodological shortcomings and risk of bias of included trials, LLLT was proved with limited evidence in delaying pain onset and reducing pain intensity. In the future, larger and better-designed RCTs will be required to provide clearer recommendations.