The imprinting process is described by a model evolved from physiological data. For the case of sexual imprinting in zebra finches a second sensitive phase, occurring at sexual maturity, is considered. The first sensitive phase is not restricted exclusively to the imprinting of sexual objects. Properties of the model allow for the explanation of apparent reversibility of imprinting. The same physiological mechanisms involved in the organization of simple neuronal connections are assumed to operate in the highly complex networks associated with imprinting phenomena.