After reviewing Fisher's correspondence relating the Ising and dimer models, we shall mention local statistics result obtained in collaboration with C. Boutillier. The latter suggest a relation between Ising configurations and spanning trees. We then prove an explicit 'correspondence' between the two models.