Signal loss due to magnetization transfer (MT) from the macromolecular protons of biological tissues is an important consideration for the in vivo detection of paramagnetic chemical exchange saturation transfer (PARACEST) agents. In this study, a four-pool model is presented that is based on the modified Bloch equations and incorporates terms for the proton exchange processes that occur in biological systems in the presence of MRI-PARACEST contrast agents. The effect of the exchangeable proton chemical shift and PARACEST agent concentration are modeled in the presence of macromolecule-derived MT. Experimental validation of the model was performed at 9.4 Tesla using Eu(3+)-DOTAM-glycine (Gly)-phenylalanine (Phe) in both aqueous solution and samples containing 10% bovine serum albumin (BSA). The model was then used to measure the agent-bound-water chemical shift and the transverse relaxation time of macromolecular protons of a sample of Vero (nonhuman primate) cells labeled with Eu(3+)-DOTAM-Gly-Phe and a phantom containing mouse brain tissue and 7 mM Eu(3+)-DOTAM-Gly-Phe. In the brain tissue phantom, a chemical shift map with standard deviation (SD) < 0.7 ppm and a T(2) map with SD < 0.6 mus were obtained. The results demonstrate the feasibility of in vivo temperature measurement based on the bound-water chemical shift of Eu(3+)-DOTAM-Gly-Phe in combination with this four-pool model despite the inherent MT effect.