This paper presents a version of hylomorphism that intends to solve problems faced by contemporary hylomorphism. After showing that attempts to understand form as sets or relation of essential properties fail at taking into account the dynamic development of substances, the paper suggests another version of hylomorphism able to solve these difficulties. A functionalist version of hylomorphism is then defended: the best way to understand how form can be present throughout all the developmental stages of a substance is to understand it as a certain kind of function.