New rapid diagnostic methods are needed to identify influenza infections to improve virological surveillance usually undertaken with conventional time-consuming, complex, and even expensive laboratory methods. Another reason for using a rapid test is to avoid inappropriate therapy, particularly in children, where use of antibiotics inappropriately and high influenza-related rates of hospitalisation are described. During two winter seasons, the performance of the QuickVue Influenza test (QV) was evaluated in children under 14 presenting with influenza like illness, and compared the results with those obtained from sentinel network surveillance using standard protocols for the sample collection and the laboratory analysis by virus culture and reverse transcription-polymerase chain reaction (RT-PCR). During the first influenza season (2000/2001), 22 paediatricians collected one nose- and one throat-swab from each of the 586 children 0-6 years old recruited in the study. The QV test was carried out in the physician's office by primary care staff on the nose swab material. When compared with virus culture of the throat swab, the QV test had a sensitivity of 36.5%. In the following 2001/2002-influenza season, the performance of the QV test as a rapid laboratory screening assay was assessed. 342 children aged 0-14 years were enrolled with only one throat swab collected from each patient and sent to the laboratory where the QV, virus culture, and RT-PCR tests were performed. The results showed a better sensitivity (54.5%) of the test in comparison with virus culture and RT-PCR assays. The data indicate that rapid QV testing in the physician office setting, using these easily obtained samples, may be too insensitive to be useful for surveillance and for immediate clinical management of children presenting with influenza-like illness. Nevertheless, the QV test may be a valuable diagnostic tool if used in laboratory, as a rapid screening test.