Aging populations are at increased risk of postoperative complications. New methods to provide care for older people recovering from surgery may reduce surgery-related complications. Comprehensive geriatric assessment (CGA) has been shown to improve some outcomes for medical patients, such as enabling them to continue living at home, and has been proposed to have positive impacts for surgical patients. CGA is a coordinated, multidisciplinary collaboration that assesses the medical, psychosocial and functional capabilities and limitations of an older person, with the goal of establishing a treatment plan and long-term follow-up.To assess the effectiveness of CGA interventions compared to standard care on the postoperative outcomes of older people admitted to hospital for surgical care.We searched CENTRAL, MEDLINE, Embase, PsycINFO, CINAHL and two clinical trials registers on 13 January 2017. We also searched grey literature for additional citations.Randomized trials of people undergoing surgery aged 65 years and over comparing CGA with usual surgical care and reporting any of our primary (mortality and discharge to an increased level of care) or secondary (length of stay, re-admission, total cost and postoperative complication) outcomes. We excluded studies if the participants did not receive a complete CGA, did not undergo surgery, and if the study recruited participants aged less than 65 years or from a setting other than an acute care hospital.Two review authors independently screened, assessed risk of bias, extracted data and assessed certainty of evidence from identified articles. We expressed dichotomous treatment effects as risk ratio (RR) with 95% confidence intervals and continuous outcomes as mean difference (MD).We included eight randomised trials, seven recruited people recovering from a hip fracture (N = 1583) and one elective surgical oncology trial (N = 260), conducted in North America and Europe. For two trials CGA was done pre-operatively and postoperatively for the remaining. Six trials had adequate randomization, five had low risk of performance bias and four had low risk of detection bias. Blinding of participants was not possible. All eight trials had low attrition rates and seven reported all expected outcomes.CGA probably reduces mortality in older people with hip fracture (RR 0.85, 95% CI 0.68 to 1.05; 5 trials, 1316 participants, I² = 0%; moderate-certainty evidence). The intervention reduces discharge to an increased level of care (RR 0.71, 95% CI 0.55 to 0.92; 5 trials, 941 participants, I² = 0%; high-certainty evidence).Length of stay was highly heterogeneous, with mean difference between participants allocated to the intervention and the control groups ranging between -12.8 and 8.3 days. CGA probably leads to slightly reduced length of stay (4 trials, 841 participants, moderate-certainty evidence). The intervention probably makes little or no difference in re-admission rates (RR 1.00, 95% CI 0.76 to 1.32; 3 trials, 741 participants, I² = 37%; moderate-certainty evidence).CGA probably slightly reduces total cost (1 trial, 397 participants, moderate-certainty evidence). The intervention may make little or no difference for major postoperative complications (2 trials, 579 participants, low-certainty evidence) and delirium rates (RR 0.75, 95% CI 0.60 to 0.94, 3 trials, 705 participants, I² = 0%; low-certainty evidence).There is evidence that CGA can improve outcomes in people with hip fracture. There are not enough studies to determine when CGA is most effective in relation to surgical intervention or if CGA is effective in surgical patients presenting with conditions other than hip fracture.