From the point-of-view of integrative biology, the question arises whether those events could be detected and followed in-vivo in an intact organism by revealing the corresponding proteolytic activity. Furthermore detecting an enzymatic activity offers the possibility of signal amplification via a renewable substrate. The first evidence that protease imaging is a pertinent way to study diseases in vivo was given using near infrared fluorimetry on a murine tumor model. Self quenched fluorescent peptides were actually cleaved and were able to generate a detectable signal in the tumor environment. In spite of recent progresses optical methods have strong limitations due to the light transmission to and from deep-seated organs. With this respect Magnetic Resonance Imaging constitutes a good alternative. Interesting IPI3063 protease and glycosidase substrates acting as MRI contrast agents have been proposed. However lower toxicity and much higher contrasts are needed to compensate for the low sensitivity of nuclear magnetic resonance. Overhauser Magnetic Resonance Imaging has the potential to significantly enhance the sensitivity of MRI. It is a double resonance experiment that transfers a fraction of the higher magnetization of the electron of a free radical to the protons of surrounding water molecules. Recently OMRI was successfully applied to in vivo oxymetry imaging by correlating the Electron Paramagnetic Resonance line width variation of a trityl free radical to oxygen concentration. Nitroxides are a family of stable free radicals. Several biocompatible nitroxides have been used in EPR and OMRI experiments in vivo. The Overhauser enhancement strictly depends on the nitroxide EPR line width. Due to the nitroxide asymmetric structure their EPR spectra significantly widen and flatten as their rotational correlation times increase. Here this property was applied to design a contrast agent sensitive to proteolysis and detectable through Overhauser enhancement. In this paper a general molecular imaging method with generation of high positive contrast in the presence of proteolytic activity is proposed. As an example nitroxides were covalently bound to bovine serum albumin. An experimental setup for OMRI at 0.2 T was built so that dynamic nuclear polarization can occur at depth in the range of a centimeter without any significant heating of the sample. OMRI of the nitroxide-labeled BSA sample PF-06651600 revealed no signal enhancement because of the elevation of the nitroxide correlation time. The initial magnetic resonance image intensity was strongly enhanced by enzymatic digestion of the carrier protein. Such a protease-switch method can be adapted to any proteolytic activity by linking a nitroxide to a large carrier molecule through specifically cleavable peptide substrates. In this paper a non invasive method designed to perform MRI of the proteolytic activity in deep-seated organs is described. This method needs a magnetic resonance imaging system including a resonant cavity tuned on the free electron EPR frequency of a nitroxide. With a very simple biochemical model, it is demonstrated that a proteolytic enzyme can modulate the Overhauser effect through the alteration of the motional correlation time of a nitroxide labelled substrate.