Patients with head and neck cancer are often malnourished and have a high incidence of postoperative complications. Studies of patients with head and neck cancer receiving immunonutrition in the perioperative period have suggested, but not conclusively demonstrated, benefit. This study reviews randomised trials comparing perioperative standard polymeric nutrition or no nutritional supplementation with immunonutrition in the treatment of head and neck cancer. Electronic databases were searched; reference lists checked and letters sent requesting details of further data. Data were combined to estimate the common relative risk of postoperative complications (wound infections, fistula formation, death and length of hospital stay), and associated 95% confidence intervals. Random effects models were used. 10 trials of polymeric nutritional supplementation with immunonutrition were identified; one compared two types of immunonutrition. There was little evidence of heterogeneity. Pooled estimates showed a reduction in length of hospital stay by 3.5 days (95% CI 0.7 to 6.3 day, P<0.01). No reductions in clinical complications were seen. Perioperative immunonutrition is associated with reduced length of hospital stay; the mechanism is unclear as other outcomes were not improved. Trials were small with incomplete reporting of outcomes. An adequately powered trial is required to substantiate benefit.