Early detection of high-grade malignancy, such as glioblastoma (GBM), using new contrast mechanism and enhanced magnetic resonance imaging (MRI) techniques increases the treatment options available, and therefore may expand survival in GBM beyond the natural history of the disease and possibly increase the patients’ survival rate. To achieve early detection, it is important to evaluate commonly used MRI techniques, as well as to develop a theoretical model of GBM growth to predict the consequential changes in MR parameters and to evaluate the performance of newly developed MRI techniques. For this purpose, popular MRI methods such as the spin-echo, the Carr-Purcell-Meiboom-Gill (CPMG), and the spin-locking radio-frequency (RF) pulse sequences were used to acquire in vivo T2-weighted and T1-weighted MR images of orthotopic GBM mouse/rat models from U87MG and C6 cell lines. Statistical results (N = 18) showed that, while the spin-echo T2-weighted MR imaging may not provide the required contrast to detect early GBM, the CPMG T2-weighted MR imaging and the spin-locking T1-weighted MR imaging did slightly improve the early GBM contrast, but at the cost of significantly elevated specific absorption rate (SAR) and increased potential risk, if applied to human patients. To facilitate the development of innovative MRI RF pulse sequences to enhance early GBM contrast with low SAR, a simple, computationally efficient theoretical model on early GBM was proposed based on the experimental observation that vessel density decreases together with the increase of vessel size during the early stage of the GBM growth. Based on this model, Monte Carlo spin dynamics simulations were carried out by solving the Bloch equations for the water magnetizations diffusing in the magnetic-field gradients induced by paramagnetic deoxyhemoglobin in the vessels. The interplay among the spin dynamics, early GBM contrast, RF pulse sequences, vessel distributions, and GBM staging can be numerically evaluated. The simulated early GBM contrast and the relaxation time constants, T2 and T1, from the spin-echo, the CPMG, and the spin-locking RF pulse sequences were compared with those from the in vivo MR imaging of orthotopic early GBM mouse models to study the relaxation mechanisms. It may also serve as a computationally efficient alternative for laboratory use of animals following the 3R (reduction, refinement and replacement) strategy in evaluating the performance of commonly use or newly developed MR methods for early GBM detection.