In this paper we intend to propose a definition of function in biology. First we delimitated the problem and pointed out that a definition must not be related to hypotheses such as the causal connexion between form and function, phylogenetics or evolution, but that only the existence of functions is to be considered. The elements to be expressed in the definition are: the relation of function to form, the dynamic character, the integration into the activity of the total organism, the possibility of latent functions and the fact that in the case of mechanical functions the function is not itself action, but directly related to it. So we proposed the following definition: A biological function is a property of a definite part of the total organism, marked by a certain form or structure, which property is actual or potential activity itself or directly related to activity, and by which the development and the maintenance of the organism as a whole is enabled to proceed in a harmonious way.