Administration of ethyl pyruvate, which is formed from pyruvate and ethanol, has been found capable of rescuing cells injured by oxidative stress. In one perspective the rescue has been postulated to be metabolic, with the resulting intracellular delivery of pyruvate seen as providing substrate for the TCA Cycle, making it possible to counteract sequela of poly(ADP-ribose)ribosylation, such as depletion of cytosolic NAD(+), glycolytic arrest, and mitochondrial deprivation of pyruvate. The rescue has also been attributed to radical scavenging via the carbonyl groups in ethyl pyruvate and pyruvate. In a previous study we exposed superfused neonatal (P7) brain slices for 60min to 2mM H(2)O(2) and found evidence for both rescue mechanisms. To see if ethyl pyruvate's actions stemmed more from being an antioxidant than from being a nutrient we conducted six new experiments using the same H(2)O(2) protocol, but with two new rescue solutions: [10mM] glucose (glc) plus one of the following: ethyl pyruvate [20mM], or the nonmetabolizable radical scavenger N-tert-butyl-alpha-phenylnitrone (PBN, 1mM). Final ATP values compared to initial, measured in 14.1T (31)P NMR spectra of PCA extracts, were the same: 0.70+/-0.08 for the former (N=3), and 0.64+/-0.08 for the latter (N=3). Quantifications of this study's (1)H NMR metabolites, also measured at 14.1T, exhibited separate clustering when pooled with data from the previous study and compared in a metabolomic multivariate analyses. Because the addition of ethyl pyruvate provided the same ATP protection as the addition of a nonmetabolizable antioxidant, antioxidant protection was its prominent protective mechanism in the chosen, high glucose protocol. Having distinct clusters in the Scores Plot of a Partial Least Squares-Discriminant Analysis suggests the feasibility of constructing statistical models that are predictive.