BACKGROUND: Although influenza is a commonly encountered condition in primary care, and diagnosis is increasingly important given the availability of new treatments, there has been no systematic review of the evidence on clinical diagnosis. METHODS: This was a systematic review of the literature with meta-analysis where appropriate. We included cohort studies and randomized trials that compared the history and physical examination with a reference laboratory test for the diagnosis of influenza A and/or B. The primary outcomes were the sensitivity, specificity, likelihood ratios, and area under the receiver-operating characteristic (ROC) curve. RESULTS: Seven studies reported the sensitivity and specificity for a total of 59 variables. We combined studies of influenza A or B alone with those of influenza A and B. Rigors [likelihood ratio (LR) +7.2], the combination of fever and presenting within 3 days of the onset of illness (LR +4.0), and sweating (LR +3.0) were best at ruling-in influenza when present. When absent, the following decreased the likelihood of influenza: any systemic symptoms (LR -0.36), coughing (LR -0.38), not being able to cope with daily activities (LR -0.39), and being confined to bed (LR -0.50). Cough, nasal congestion, and fever (subjective or objective) had the highest calculable areas under the ROC curve. CONCLUSIONS: Individual signs and symptoms are of limited value for the diagnosis of influenza. Development of clinical decision rules that systematically combine symptoms may be a more useful strategy.