Background Indications for the use of negative pressure wound therapy (NPWT) are broad and include prophylaxis for surgical site infections (SSIs). While existing evidence for the effectiveness of NPWT remains uncertain, new trials necessitated an updated review of the evidence for the effects of NPWT on postoperative wounds healing by primary closure. Objectives To assess the effects of negative pressure wound therapy for preventing surgical site infection in wounds healing through primary closure. Search methods We searched the Cochrane Wounds Specialised Register, CENTRAL, Ovid MEDLINE (including In‐Process & Other Non‐Indexed Citations), Ovid Embase, and EBSCO CINAHL Plus in February 2018. We also searched clinical trials registries for ongoing and unpublished studies, and checked reference lists of relevant included studies as well as reviews, meta‐analyses, and health technology reports to identify additional studies. There were no restrictions on language, publication date, or setting. Selection criteria We included trials if they allocated participants to treatment randomly and compared NPWT with any other type of wound dressing, or compared one type of NPWT with another type of NPWT. Data collection and analysis Four review authors independently assessed trials using predetermined inclusion criteria. We carried out data extraction, 'Risk of bias' assessment using the Cochrane 'Risk of bias' tool, and quality assessment according to GRADE methodology. Main results In this second update we added 25 intervention trials, resulting in a total of 30 intervention trials (2957 participants), and two economic studies nested in trials. Surgeries included abdominal and colorectal (n = 5); caesarean section (n = 5); knee or hip arthroplasties (n = 5); groin surgery (n = 5); fractures (n = 5); laparotomy (n = 1); vascular surgery (n = 1); sternotomy (n = 1); breast reduction mammoplasty (n = 1); and mixed (n = 1). In three key domains four studies were at low risk of bias; six studies were at high risk of bias; and 20 studies were at unclear risk of bias. We judged the evidence to be of low or very low certainty for all outcomes, downgrading the level of the evidence on the basis of risk of bias and imprecision. Primary outcomes Three studies reported mortality (416 participants; follow‐up 30 to 90 days or unspecified). It is uncertain whether NPWT has an impact on risk of death compared with standard dressings (risk ratio (RR) 0.63, 95% confidence interval (CI) 0.25 to 1.56; very low‐certainty evidence, downgraded once for serious risk of bias and twice for very serious imprecision). Twenty‐five studies reported on SSI. The evidence from 23 studies (2533 participants; 2547 wounds; follow‐up 30 days to 12 months or unspecified) showed that NPWT may reduce the rate of SSIs (RR 0.67, 95% CI 0.53 to 0.85; low‐certainty evidence, downgraded twice for very serious risk of bias). Fourteen studies reported dehiscence. We combined results from 12 studies (1507 wounds; 1475 participants; follow‐up 30 days to an average of 113 days or unspecified) that compared NPWT with standard dressings. It is uncertain whether NPWT reduces the risk of wound dehiscence compared with standard dressings (RR 0.80, 95% CI 0.55 to 1.18; very low‐certainty evidence, downgraded twice for very serious risk of bias and once for serious imprecision). Secondary outcomes We are uncertain whether NPWT increases or decreases reoperation rates when compared with a standard dressing (RR 1.09, 95% CI 0.73 to 1.63; 6 trials; 1021 participants; very low‐certainty evidence, downgraded for very serious risk of bias and serious imprecision) or if there is any clinical benefit associated with NPWT for reducing wound‐related readmission to hospital within 30 days (RR 0.86, 95% CI 0.47 to 1.57; 7 studies; 1271 participants; very low‐certainty evidence, downgraded for very serious risk of bias and serious imprecision). It is also uncertain whether NPWT reduces incidence of seroma compared with standard dressings (RR 0.67, 95% CI 0.45 to 1.00; 6 studies; 568 participants; very low‐certainty evidence, downgraded twice for very serious risk of bias and once for serious imprecision). It is uncertain if NPWT reduces or increases the risk of haematoma when compared with a standard dressing (RR 1.05, 95% CI 0.32 to 3.42; 6 trials; 831 participants; very low‐certainty evidence, downgraded twice for very serious risk of bias and twice for very serious imprecision. It is uncertain if there is a higher risk of developing blisters when NPWT is compared with a standard dressing (RR 6.64, 95% CI 3.16 to 13.95; 6 studies; 597 participants; very low‐certainty evidence, downgraded twice for very serious risk of bias and twice for very serious imprecision). Quality of life was not reported separately by group but was used in two economic evaluations to calculate quality‐adjusted life years (QALYs). There was no clear difference in incremental QALYs for NPWT relative to standard dressing when results from the two trials were combined (mean difference 0.00, 95% CI −0.00 to 0.00; moderate‐certainty evidence). One trial concluded that NPWT may be more cost‐effective than standard care, estimating an incremental cost‐effectiveness ratio (ICER) value of GBP 20.65 per QALY gained. A second cost‐effectiveness study estimated that when compared with standard dressings NPWT was cost saving and improved QALYs. We rated the overall quality of the reports as very good; we did not grade the evidence beyond this as it was based on modelling assumptions. Authors' conclusions Despite the addition of 25 trials, results are consistent with our earlier review, with the evidence judged to be of low or very low certainty for all outcomes. Consequently, uncertainty remains about whether NPWT compared with a standard dressing reduces or increases the incidence of important outcomes such as mortality, dehiscence, seroma, or if it increases costs. Given the cost and widespread use of NPWT for SSI prophylaxis, there is an urgent need for larger, well‐designed and well‐conducted trials to evaluate the effects of newer NPWT products designed for use on clean, closed surgical incisions. Such trials should initially focus on wounds that may be difficult to heal, such as sternal wounds or incisions on obese patients.