Let B be an n × n real expanding matrix and D be a finite subset of R n with 0 ∈ D. The self-affine set K = K(B, D) is the unique compact set satisfying the setvalued equation BK = d∈D (K + d). In the case where card(D) = |det B|, we relate the Lebesgue measure of K(B, D) to the upper Beurling density of the associated measure µ = lim s→∞ ℓ0,...,ℓs−1∈D δ ℓ0+Bℓ1+···+B s−1 ℓs−1 . If, on the other hand, card(D) < |det B| and B is a similarity matrix, we relate the Hausdorff measure H s (K), where s is the similarity dimension of K, to a corresponding notion of upper density for the measure µ.