This paper proposes an application of some particular theories, known as the ‘polysystem’ theories of translation, to the study of film adaptation. A preliminary and experimental analysis of a series of film adaptations made in the American film noir of the 1940s and 1950s shows that this approach provides the basis for a systematic and coherent method with theoretical foundations, and that it permits the study of aspects of film adaptation which have been neglected or ignored so far.