The extensive laboratory investigations of respiratory disease in the U.K. over many years have demonstrated the frequency with which influenza viruses, both A and B, are found each winter. Only rarely are none isolated. These findings correlate well with other indicators of influenza such as increases in sickness benefit claims and in deaths attributed to influenza and pneumonia. However, outside these demonstrable peaks of incidence influenza viruses have been found to circulate over considerably longer periods often first appearing as early as November and continuing through to April or even May. But there has been no regular or predictable pattern determined. The period of 1968-76 has seen a series of differently developing influenza winter epidemics caused by a series of the H3N2 virus. The contributions of virus isolation and serology to influenza surveillance is discussed.