Human epidemiology differs from studies on health care. Epidemiology is an observational type of research, and partially a substitute for experimental investigations which are seldom possible in the human situation. Another difference from the human situation are privacy features both on the level of individuals and on that of organisations. This is in contrast with herd books, artificial insemination banks and slaughterhouse data bases in animals. In view of these differences and limitations, the largest potential for veterinary epidemiology would appear to lie in the fields of data and biological banking. Instances of the routine collection of data to monitor disease in human individuals are presented. The use of biological banks is discussed: blood, urine, nails, etc., are collected from a cohort of (still) normal animals in this case. This is done in order to use them after follow-ups to determine pre-diagnostic levels and changes in biological parameters in retrospect. This approach, consisting in case-control studies within a cohort combines the advantages of the two methods of investigation, namely the longitudinal aspect which makes it possible to answer post-propter questions and the cost-saving aspect of confining biochemical analysis to cases alone and the controls proper to these cases. Drawbacks to the two separate techniques, such as an expensive complete follow-up and selection bias problems, are ruled out in this manner.