Abstract. After the introduction in 1994, by Okabe and Matsuda, of the notion of semistar operation, many authors have investigated different aspects of this general and powerful concept. A natural development of the recent work in this area leads to investigate the concept of invertibility in the semistar setting. In this paper, we will show the existence of a "theoretical obstruction" for extending many results, proved for star-invertibility, to the semistar case. For this reason, we will introduce two distinct notions of invertibility in the semistar setting (called ⋆-invertibility and quasi-⋆-invertibility), we will discuss the motivations of these "two levels" of invertibility and we will extend, accordingly, many classical results proved for the d-, v-, t-and w-invertibility. Among the main properties proved here, we mention the following: (a) several characterizations of ⋆-invertibility and quasi-⋆-invertibility and necessary and sufficient conditions for the equivalence of these two notions; (b) the relations between the ⋆-invertibility (or quasi-⋆-invertibility) and the invertibility (or quasi-invertibility) with respect to the semistar operation of finite type (denoted by ⋆ f ) and to the stable semistar operation of finite type (denoted by ⋆), canonically associated to ⋆; (c) a characterization of the H(⋆)-domains in terms of semistar-invertibility (note that the H(⋆)-domains generalize, in the semistar setting, the H-domains introduced by Glaz and Vasconcelos); (d) for a semistar operation of finite type a nonzero finitely generated (fractional) ideal I is ⋆-invertible (or, equivalently, quasi-⋆-invertible, in the stable semistar case) if and only if its extension to the Nagata semistar ring I Na(D, ⋆) is an invertible ideal of Na(D, ⋆).