Spatially resolved investigations of molecular properties in polymer materials are shown to be possible by means of parameter-selective NMR imaging. The transformation of NMR parameters into material properties can be performed using a suitable theoretical description of the molecular mobility in the polymer under investigation. Rubber samples consisting of three pieces of rubber with different cross-link densities have been investigated in a first experiment. A series of images which were progressively weighted by transverse relaxation have been acquired by application of a spin-echo back-projection technique. The relaxation was analyzed pixel by pixel using a single chain model for the molecular dynamics of the polymer network. Parameters directly related to material parameters, such as cross-link density and correlation times, could be derived from the theory and were displayed in the form of a MAterial Property image (MAP image). In contrast to conventional spin density images, where the different pieces of rubber could not be distinguished properly, the MAP image clearly reflects the different cross-link densities. Moreover, histogram analysis of the image can be used for the characterization of cross-link density inhomogeneities.