The potential usefulness of computer-aided diagnosis (CAD) in mass survey for lung cancer were investigated. When the sensitivity of the computer output was adjusted nearly equal to that of the mass survey in our database which contained 198 photoflurofaraphic films, some shadows detected by the computer output were different from those detected by human observers. Therefore, the best estimated sensitivity of an observer using the computer output was equal to or greater than the sensitivity of double reading. It is expected that CAD may have a role in mass survey for lung cancer using photofluorographic films.