Abstract. Every new programming technique makes claims that software engineers want to hear. Such is the case with aspect-oriented programming (AOP). This paper describes a quasi-controlled experiment which compares the evolution of two functionally equivalent programs, developed in two different paradigms. The aim of the study is to explore the claims that software developed with aspect-oriented languages is easier to maintain and reuse than this developed with object-oriented languages. We have found no evidence to support these claims.