Various aspects of the British statistics relating to sexually-transmitted diseases are examined. Defects in these are illustrated by data from a series of surveys. It is suggested that the information being collected and published is incomplete by an unknown amount, inadequate because it tells only of laboratory diagnoses, and unusable at a local level because it relates to no identifiable population. Some suggestions are made as to how the data collected might be made more useful: the first and most important is that a working party should decide for what purposes the statistics are required.