We study the scaling limit of the capacity of the range of a simple random walk on the integer lattice in dimension four. We establish a strong law of large numbers and a central limit theorem with a non-gaussian limit. The asymptotic behaviour is analogous to that found by Le Gall in '86 [28] for the volume of the range in dimension two.One easily passes from one representation in (1.1) to the other using the last passage decomposition formula, see (2.8) below. Denote by {S(n), n ∈ N} a simple random walk in Z 4 . For two integers m, n, the range R[m, n] (or simply R n when m = 0) in the time period [m, n] is defined as R[m, n] = {S(m), . . . , S(n)}.Our first result is a strong law of large numbers for Cap (R n ).Recently, van den Berg, Bolthausen and den Hollander [38] advocated a new geometric characteristic, the torsional rigidity of the complement of a Wiener sausage, as a way to probe the shape of the sausage. In order to obtain leading asymptotics for the torsional rigidity, one needs a law of large numbers for the capacity of a Wiener sausage, which is not proved yet in dimension four; see however our companion paper [5] for a partial result in this direction. Our Theorem 1.1 establishes these asymptotics for the discrete model, and thus prepares the study of torsional rigidity for random walk.