Activators of the transcriptional factor, Nrf-2 have been shown to diminish cell dysfunction and death secondary to a host of nervous system insults in vitro and in vivo. It is believed that Nrf-2 mediates its salutary effects in part, via the upregulation of genes that counteract oxidative stress in astrocytes, with concordant cell autonomous and non cell autonomous protective effects in the CNS. However, the upstream activators of Nrf-2 remain obscure. To address this important question, we developed a novel method to spatially and temporally manipulate levels of an important reactive oxygen species, hydrogen peroxide in mixed astrocyte-neuronal cultures. The method involves the heterologous expression of d-amino acid oxidase in astrocytes not neurons, followed by addition of the d-amino acid, d-alanine. Low levels of d-alanine addition (16 micromolar) lead to generation of 3.7 nM of peroxide per min/mg of protein only in astrocytes not untransfected neurons. 16 micromolar d-alanine for seven hours induces a state change in astrocytes that permits protection of adjacent neurons from oxidative death (Haskew-Layton et al., 2010). By contrast, higher concentrations of d-alanine (2 millimolar) lead to 130 nM/min peroxide produced/mg protein and toxicity of adjacent neurons. Unexpectedly, protective concentrations of peroxide neither activate Nrf-2 nor require Nrf-2 for their salutary effects; rather it appears that low peroxide can lead to inactivation of the phosphatase PTEN leading to modulation of gene expression including the enzyme transglutaminase 2 (TG2). Indeed, TG2 inhibition in astrocytes leads to protection of neurons from oxidative death (Basso et al., 2012). These results demonstrate the importance of understanding the spatial and temporal regulation of peroxide in astrocytes, neurons and other cells of the CNS; and suggest that peroxide is not a dominant mediator of Nrf-2 activation. While peroxide induces a protective “state” in astrocytes, the magnitude of the protection remains inferior to canonical activators of Nrf-2, including sulforafane and tert-butyl hydroquinone. However, a theoretical limitation in the application of canonical activators of the Nrf-2 pathway to neurological disease lies in the fact that most of these compounds are “electrophiles”. Under basal conditions, these electrophiles alkylate the cytoplasmic inhibitor Keap to activate Nrf-2; however, under pathological conditions where redox balance is disrupted, electrophiles would be more likely induce toxicity. To overcome this theoretical problem, we have developed a novel screening and monitoring tool to enable us to identify non toxic activators of the Nrf-2 pathway. Specifically, we fused the Neh2 domain of Nrf-2 to luciferase. The Neh2 domain is the putative negative regulatory domain that interacts with Keap1. The Neh2-luciferase construct was more sensitive with a wider dynamic range than currently available reporters for Nrf-2; further it is induced by molecular knockdown of Keap1 and suppressed by overexpresion of Keap1 (Smirnova et al., 2011). We screened a library of 2000 compounds from the Microsource Library and identified a number of compounds that activated our reporter; secondary assays confirmed that these hits could active Nrf-2 dependent gene expression in primary neurons. Based on kinetic studies, and assumption of equal bioavailability of our compounds, we have sub classified the activators into five groups. Among the groups identified, those activated via a switch or receptors were deemed to be the most effective and least toxic. Among this group, nordihydroguairetic acid, fisetin and gedunin represented the best hits in the screen. These hits activated Nrf-2 dependent gene expression in astrocytes or neurons and protected these cells from oxidative stress in an Nrf-2 dependent manner. Overall, our studies identify novel and safer activators of Nrf-2 that induce neuroprotection.