Screening for an early detection of disease, particularly when the screening tests have to be carried out on a number of occasions during the lifetime of a person, is a complex issue and the development of suitable models that can help in setting up good screening programmes for a community is a challenging task. In most cases the current emphasis is on screening at regular intervals of time and the frequency of screening is determined, usually in an informal way, by factors like the nature of the test and the nature of the disease. Analysis of a simple two-state model leads to a basic building-block which can be used in the construction of models for a variety of screening programmes. The use of this building block is illustrated and it is shown that the determination of non-regular screening times in a formal manner can lead to a greater efficiency.