This review covers two conditions: acute clavicle fractures and non-union resulting from failed fracture healing. Clavicle (collarbone) fractures account for around 4% of all fractures. While treatment for these fractures is usually non-surgical, some types of clavicular fractures, as well as non-union of the middle third of the clavicle, are often treated surgically. This is an update of a Cochrane review first published in 2009. To evaluate the effects (benefits and harms) of different methods of surgical treatment for acute fracture or non-union of the middle third of the clavicle. We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (27 June 2014), the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2014, Issue 5), MEDLINE (1966 to June week 3 2014), EMBASE (1988 to 2014 week 25), LILACS (1982 to 27 June 2014), trial registries and reference lists of articles. We applied no language or publication restrictions. We considered randomised and quasi-randomised controlled trials evaluating any surgical intervention for treating people with fractures or non-union of the middle third of the clavicle. The primary outcomes were shoulder function or disability, pain and treatment failure (measured by the number of participants who had undergone or were being considered for a non-routine secondary surgical intervention for symptomatic non-union, malunion or other complications). Two review authors selected eligible trials, independently assessed risk of bias and cross-checked data. Where appropriate, we pooled results of comparable trials. We included seven trials in this review with 398 participants. Four trials were new in this update.The four new trials (160 participants) compared intramedullary fixation with open reduction and internal fixation with plate for treating acute middle third clavicle fractures in adults. Low quality evidence from the four trials indicated that intramedullary fixation did not result in a clinically important improvement in upper arm function (despite a statistically significant difference in its favour: standardised mean difference 0.45, 95% confidence interval (CI) 0.08 to 0.81; 120 participants, three trials) at long term follow-up of six months or more. Very low quality evidence indicated little difference between intramedullary fixation and plate fixation in pain (one trial), treatment failure resulting in non-routine surgery (2/68 with intramedullary fixation vs. 3/65 with plate fixation; risk ratio 0.69, 95% CI 0.16 to 2.97, four trials) or time to clinical fracture consolidation (three trials). There was very low quality evidence of a lower incidence of participants with adverse events (mainly infection, poor cosmetic result and symptomatic hardware) in the intramedullary fixation group (18/68 with intramedullary fixation vs. 27/65 with plate fixation; RR 0.64, 95% CI 0.39 to 1.03) but the CI of the pooled results also included the small possibility of a lower incidence in the plate fixation group. None of the four trials reported on quality of life or return to previous activities. Evidence is pending from two ongoing trials, with planned recruitment of 245 participants, testing this comparison.There was low or very low quality evidence from three small trials, each testing a different comparison. The three trials had design features that carried a high risk of bias, potentially limiting the reliability of their findings. Low-contact dynamic compression plates appeared to be associated with significantly better upper-limb function throughout the year following surgery, earlier fracture union and return to work, and a reduced incidence of implant-associated symptoms when compared with a standard dynamic compression plate in 36 adults with symptomatic non-union of the middle third of the clavicle. One quasi-randomised trial (69 participants) compared Knowles pin versus a plate for treating middle third clavicle fractures or non-union. Knowles pins appeared to be associated with lower pain levels and use of postoperative analgesics and a reduced incidence of implant-associated symptoms. One study (133 participants) found that a three-dimensional technique for fixation with a reconstruction plate was associated with a significantly lower incidence of symptomatic delayed union than a standard superior position surgical approach. Evidence is pending from two ongoing trials, with planned recruitment of 130 participants, comparing anterior versus superior plates for acute fractures. There is very limited and low quality evidence available from randomised controlled trials regarding the effectiveness of different methods of surgical fixation of fractures and non-union of the middle third of the clavicle. The evidence from four ongoing trials is likely to inform practice for the comparisons of intramedullary versus plate fixation and anterior versus superior plates for acute fractures in a future update. Further randomised trials are warranted, but in order to optimise research effort, these should be preceded by research that aims to identify priority questions.