Amyotrophic lateral sclerosis (ALS) is a devastating disease characterized by the progressive degeneration of motor neurons. Dominantly-inherited mutations to the antioxidant enzyme Cu, Zn superoxide dismutase (SOD1) cause 3-6% of all ALS cases. The complete mechanism behind the toxicity of mutant SOD1 remains unclear, although significant evidence points to aberrant or incomplete metal-binding having a role in a toxic gain-of-function. However, the relevance of the metal-binding of SOD1 to mutant-SOD1-linked ALS remains controversial. Direct assessments of protein metal-binding from transgenic, SOD1-overexpressing rodent models of the disease are difficult to acquire due to the non-covalent nature of the interaction. The relatively small amount of disease-afflicted spinal cord tissue in which the motor neurons reside compounds the difficulty of measuring the protein metal binding of SOD1 from transgenic mice. This dissertation addresses the metals bound to SOD1 throughout the disease course in transgenic mice using a novel mass spectrometry assay. The methodology developed here offers the first detailed examination of partially unfolded intermediates of SOD1 present in the spinal cord of pre-symptomatic, symptomatic, and end-stage transgenic mice overexpressing the ALS-associated SOD1 mutation G93A (glycine mutated to alanine at position 93). These results were compared to age-matched transgenic mice expressing wild-type SOD1 that do not develop ALS symptoms. To extract SOD1 from relevant spinal cord tissue, a 300 [micro]m necropsy punch was used to remove a small piece of tissue from the ventral or dorsal gray matter of a 1 mm-thick slice of spinal cord. Physiological salts that interfere with electrospray mass spectrometry were removed by binding the proteins to a C4 Ziptip®, a pipette tip containing hydrophobic, reversed-phase packing material. Washing the Ziptip-bound proteins with water eliminated interfering salts. Bound proteins could then be eluted into a mass spectrometer with low concentrations of acetonitrile plus formic acid. Electrospray ionization conditions were determined that could keep both copper and zinc bound to SOD1. Using a high-resolution Fourier transform-ion cyclotron resonance mass spectrometer, we used the assay to collect isotopically-resolved protein mass data. Theoretical protein isotope distributions were calculated from the empirical formulas of SOD1 and matched to the experimental data with a least squares fitting algorithm to determine the multiple intermediates of SOD1 present. Spinal cord tissue, wild-type in particular, was notable for containing significantly more one-metal SOD1 than any other tissue, despite having 3-fold less SOD1 than liver. We quantitatively compared the levels of soluble, partially unfolded intermediates of SOD1 from wild-type and G93A SOD1 spinal cords. Wild-type mouse spinal cord contained significantly more of all of the partially unfolded intermediates copper-deficient SOD1, disulfide reduced SOD1, and apo SOD1. The amount of zinc-containing SOD1 was exceptionally high in wild-type mice, comprising 60% of the total SOD1 in wild-type spinal cord. The larger amounts of these SOD1 intermediates in wild-type transgenic mice indicate that they are not directly responsible for toxicity in vivo. However, copper-containing, zinc-deficient SOD1 was the one species found in higher concentrations in G93A SOD1 spinal cord. The concentration was on average 0.6-0.8 [micro]M in G93A spinal cord, compared to 0.1-0.3 [micro]M zinc-deficient SOD1 found in the wild-type mouse spinal cord. A concentration above 0.5 [micro]M zinc-deficient SOD1 was sufficient to induce motor neuron death in vitro. These results suggest that copper-containing, zinc-deficient SOD1 could be the toxic species responsible for motor neuron death in ALS.