During the past decades the interest in surveillance and monitoring of wildlife diseases has grown internationally. The main reasons could be the following: a) increased size of many wild populations that host pathogens affecting humans; b) the increased economic relevance of some wildlife disease; c) the role played by infections/diseases in the conservation of some wild endangered species. According to the above-described epidemiological situations there is an international need to develop appropriate strategies for the early detection, monitoring and surveys of infectious diseases in wildlife. The paper reviews the epidemiological assumptions on which disease surveillance, monitoring and survey are, or should be, based. The main conclusions are: 1) wildlife disease surveillance and monitoring are long lasting activities that should be implemented when legal bases are available; 2) a wildlife disease introduced in a free area is more likely to be early detected using passive rather than active surveillance; 3) the definition of the "suspect case" largely affects the sensitivity of the whole passive surveillance; thus the suspected case definition should be modulated according to the level of risk; 4) in both active surveillance and monitoring, sampling plays an important role. The sensitivity of any active surveillance/monitoring system is highly dependent from the sampling unit that we define as: "the host subpopulation, whose size can maintain the pathogen during a defined inter-sampling interval". Such definition merges the ecological, epidemiological and mathematical approaches aimed in controlling or eradicating infections in both domestic and wildlife; 5) When dealing with the conservation-disease interface, a standardized risk assessment procedure including risk mitigation has to become the rule.