Abstract. In this paper the notion of two-scale convergence introduced by G. Nguetseng and G. Allaire is extended to the case of bounded sequences in L 1 (Ω), where Ω is any open subset of R N . Three different approaches will be considered: an adaptation of the method used in L p (Ω) with p > 1, a measure-theoretic argument, and the periodic unfolding technique.