Sodium (23Na) plays a pivotal role in cellular homeostasis throughout all life forms. In the human body it has a plethora of physiological functions including regulation of bodily fluids, muscle contraction, neuronal transmission, and connective tissue integrity. The 23Na nucleus can be visualized using magnetic resonance imaging (MRI) because it has spin 3/2 and is the 100% naturally abundant isotope of sodium. However, unlike hydrogen (1H), the nucleus of choice for MRI scanning, 23Na MR clinical applications are almost non-existent, even though it is the second most MR visible nucleus after 1H. This is primarily because 23Na requires its own customized hardware (i.e. radiofrequency RF coil) and software (i.e. pulse sequence/acquisition and image reconstruction techniques). Additionally, 23Na MR is challenging because of significantly lower signal-to-noise ratio (SNR), which translates to a requirement for lengthy scans and lower spatial resolution. Hence, development and optimization of 23Na MR acquisition techniques is an ongoing subject of research. In this dissertation development and optimization of approaches that exploit sodium as a biomarker for musculoskeletal health are described. Specifically, discussions of advanced 23Na MR techniques fall into two main categories: magnetic resonance imaging (MRI) and spectroscopy (MRS). The work first explored development of a MRI compatible electrical muscle stimulation (EMS) system that could be used as a consistent alternative to voluntary muscle contraction. A variety of MRI methods were used to confirm the similarity between stimulation and voluntary activation of muscles of the lower leg. In a follow- up study quantum filtered (QF) 23Na MRS was used to assess dynamic changes in skeletal muscle before, during and following voluntary exercise. Total (single quantum filtered, SQF) and bound intracellular sodium (triple quantum filtered, TQF) were measured in 9 healthy subjects with a 12s temporal resolution. Total sodium (SQF) significantly increased 4% (p < 0.01), while bound intracellular sodium content decreased 7% (p < 0.01) with exercise. Both returned to baseline following exercise; TQF after a few seconds and SQF after approximately 12 minutes (p < 0.05). In a preliminary study the MR-compatible EMS unit demonstrated similar QF 23Na MRS results. Sodium MR imaging, like typical 1H-MRI scanning, can be performed by rasterizing k-space. However, this approach is both temporally and SNR inefficient. There- fore newer optimized 23Na MRI pulse sequence approaches have been developed. The cost-benefit of extending the acquisition window length in terms of SNR gain and blurring in the in vivo 23Na MRI of the human knee was explored in 3 healthy subjects using a density adapted 3-dimensional projection reconstruction (DA-3DPR) sequence. Mean SNR doubled when the acquisition window was increased from 4 to 25ms. Concurrently, the FWHM, as a measure of cartilage blurring, increased by only 1±0mm across three different sections of articular cartilage. In a second imaging experiment, a pseudo-random 3D non-Cartesian k-space acquisition scheme for 23Na MRI was introduced. This scheme was highly effective in minimizing the aliasing artifacts leading to 9 times less number of shots required to cover k-space compared to a fully-sampled acquisition scheme. Hence, this resulted in a 9-fold reduction in scan time to cover k-space. Images of a resolution phantom and healthy human knee, with 3mm isotropic resolution (zero padded to 1mm isotropic resolution), were reconstructed using a non-uniform fast Fourier transform (NUFFT). / Dissertation / Doctor of Philosophy (PhD)