Abstract
A considerable amount of background epidemiological information is required for a rational approach to influenza vaccination. This is discussed and illustrated with tables and figures. Examples are given of the variation of the various strains and the problem of anticipation and prediction of epidemics discussed.